Justin Taylor

Justin, thanks for coming online. So as well as being a motion designer, I adore and nerd out a lot about tools for motion. I've spoken to a few tool makers here on this podcast, like Adam, Ari, Danny, some others. And like many of these tool makers, not necessarily all of them, build a lot of what they do on top of your frameworks. So I like to think of you as like the tool maker's tool maker in relation to video and motion and stuff like that. And a lot of people probably don't know that about Hyperbrew. ⁓ Presumably like the reason you entered this world is like one, you needed a better experience yourself. ⁓ but it also may be like now creates more of a pipeline for your other, the other side of your business, which is like custom development, or maybe that's the main side. So I guess maybe you could just like go back and give us a broad overview of like how you got started in this and then how you ended up making kind of frameworks and tools for motion design tooling.

Yeah, absolutely. So it really started out a necessity. You know, in my first job, I was shooting and editing full time, doing a lot of product videos. And my day to day job was super repetitive. We had like dozens of product videos we had to churn out per week. And so there was a lot of like, you know, open the product, put it on the table, film it, go to the next one. And then when it comes to Premiere Pro, it's like we're editing the exact same templates over and over again. So, you know, after doing that for a while to make my own job more interesting and try to focus a little more on the parts of my job that I enjoy rather than the mundane repetitive parts. I was like, hey, you know, can we automate some of this stuff? What's possible? So that's when I kind of started looking into ⁓ CEP extensions for Premiere Pro, is, doesn't have script UI like After Effects does, but it has CEP extensions, which are like a web browser on the front end ⁓ and JavaScript on the back end. ⁓ So started with that, was pretty new to JavaScript at the time, but learned just enough to start automating some of those steps in own workflow. ⁓ So that's kind of how it started was just like for myself because I'm like okay now I don't have to set up the project the same way every time I can click a button it'll go to the right folders import the footage for me and when it comes to export I don't yeah sure yeah

And without moving on too fast from that, I just, cause, cause yeah, I did ask a big question. Like I'm interested in like one, like the liftoff time for you from like learning. back then we didn't have language models or anything like what we probably did. They were just a bit rubbish. I don't know. You can give us the timeline. ⁓ And ⁓ like the, just interested in like, yeah, how long it took you to be able to create something that was actually genuinely valuable and how valuable was it? Like did people on your team notice it? ⁓ Maybe what was it? ⁓ and then also a second question is like, what was the kind of development experience like? Cause I want to kind of keep these two threads here of like, yeah, so we can end up in the right place.

Totally. ⁓ Yeah, development experience was not great. And I think there's a lot of, you There's a lot of discussion online about how hard it is to get started building tools for Adobe apps. The challenges with, whether it's lack of documentation or resources and outdated references and that sort of thing. ⁓ But I mean, how long did it take me to get something up and running that was valuable? It didn't take too long actually, I could say that. The first version of a tool that was useful to me, it looked totally ugly. The code was terrible, but it saved time for me. in my job. And so for that time, was a small team, it was just really me, two other guys in the photo video department. So it was really just for myself, but I would show them and be like, hey, look at what I built, you know, and stuff. It was mainly just useful for me at the time. It wasn't until my second job at Verasity CoLab that we had more of a mid-size video team. I kind of started sharing more with other motion designers and video people. And then I started to get more input and developing it further and stuff. But yeah, I would say for that first version of something that was useful, like ⁓ Gosh, probably a couple weeks, you know, and that was like with minimal, you know, programming experience and stuff in the past, probably, I mean, I some like web dev stuff in the past, just like basic HTML, CSS, and like a tiny bit of JavaScript, but not too much. So to get something up and running, yeah.

Mm-hmm. And was there enough context online that you could like find out like either templates or pre-existing code or stack overflow or anything like that? Or was it really that you had to learn from scratch and make it yourself?

I mean, that kind of is learning from scratch, which is what you describe, right? When you're like piecing together all the pieces versus, you know, going through a kind of structured education when you're having to say, okay, I want to build a plugin. And so, you know, you go to Adobe's website and they're like, okay, here's a template and you download it. And you're like, cool, now what do I do? And so then you have to kind of start reading through the docs and then. where there's like parts of the docs that are confusing, going to the forums, looking at past questions, asking any questions. So yeah, there was a lot of that, that kind of piecemeal finding your way to kind of get to that first version of something that's working. But the first time you get something that's working, I'm sure you've experienced this, because I know you've done some script development yourself is like, ⁓ there's that satisfaction when you first get something that's like you click a button, and then it does something that's useful for you. And then you're like, ⁓ that is so nice. How can we make this better? What else can it do? And it kind of opens that door of something that you didn't know was possible. You're not like, I'm just stuck with whatever's in Premiere, After Effects, or what's on Aescripts. It's like, actually, there's possibilities that would literally be endless if I just put the time in to learn it. ⁓ So yeah, that's kind of how it started. Sure. Mm-hmm.

Mm-hmm. Mm-hmm. Yeah. Yeah. And I don't want to move away from Verasity and like, I think you worked at some other places, other places before you did Hyper-U as well. So we won't, we'll come back to those, but like, yeah, it is that thing. I think maybe we spoke about it offline about like this thing about like you're doing the thing, right? The motion, the video, whatever. And then you're like, I need a thing to do the thing. You go over and do this thing. And you're like, this thing's really interesting. I'm going to go back and do the thing. And then you, Adam, other people, myself, I'm like kind of, I feel like I'm halfway between this. like, I'm still employed to do video.

Sure.

And I'm in this place where I'm like, I absolutely love making motion tools, but like, is a somewhat of a distraction. It speeds us up. have been like making some stuff recently, which is quite helpful for our team. But ⁓ yeah, that must've been an interesting decision. So at that point, were you thinking like, I want to do this or was it still just like, this is helping me do what I do. Like when you're at Verasti.

Yeah, I mean it was always secondary, you my job was editing and motion design and... this was just something to speed up my own workflow. then when I was working with the team, there was more people doing the same thing I did. I was like, okay, how can we make this more flexible, make the UI a little less ugly, make it more user friendly and better UI, better UX. ⁓ But yeah, it was always a secondary thing ⁓ just to make my workflow faster. And that's the case for a lot of people, technical directors or just technically minded motion designers. ⁓ Their focus is the art and they love the art, but they know some scripting or they know some ways of automating things. ⁓ And that's only gonna help you, it's only gonna help your team and stuff going forward. So it doesn't mean that you have to totally switch gears like I did and go full into the coding world. ⁓ But yeah, however you wanna take it, it's only gonna help for sure.

I'm interested as well, when, cause I assume like there's a kind of, ⁓ a huge motivation for you, like, there is for me, like when you make tools that help other people do more, be more, be more creative and seeing like the joy in them and also what they create as well. When did that start kicking in? Was that at Verasti or was it later or?

Yeah, I mean, would say Veracity because that was the first time I was able to work with other people who they could actually start using it and then be like, dude, this is awesome. Like, this just saved me time doing this or like, yeah, exactly. So yeah, and so that was cool to kind of see like, okay, this is something that's not just for me, you know, it can help other people. And then eventually getting my first tool and it he scripts and then, now it's helping the world basically and stuff. So yeah, the first tool was Proario. ⁓ And so it's, yeah. ⁓

And feature requests are like fun at that time. Well, I mean, they may still be now, but who knows. What was that first of all? wow, okay.

familiar ProIO automates the importing and exporting in After Effects and Premiere Pro. instead of having to set up all of your importing footage manually, dragging and dropping, you can set up watch folders that you can just import from automatically. And then when it comes to exporting, if you have complex exports, prioritize H.264, the list goes on. You can set up presets. It's like a one button click and you can export to Media Encoder or After Effects Render Queue and stuff. ⁓ So that was, yeah.

It's sick. I love that. That's like one of the first plugins I ever bought for After Effects. Cause I came from the music world and also the design world. And like in both of those worlds, the paradigm is not that much file linking like logic and Ableton tend to sort of manage it for you to an extent. Obviously you miss files sometimes, but logic is like just one file and it all gets chucked in there. If you want, you can change that setting, but that's the default. And then when I was just learning particularly motion and I was just working with tons of like, not just PNG sequences, but lots of UI animation with like

no way. that's awesome. Sure. Yeah. Yeah.

exported PNGs like to make it all work. I was just like, this is mental. Like I have to update that one thing and like track it and then like replace it. So that's why I found it. ⁓ it's a really interesting, like, yeah, we'll, we'll get into this, but like, think why I find the tooling around motion and video so interesting is just, there's so much work to be done still. People think like, like tools for video and, and, and motion and like they've been around forever. There's so many, it's all covered, but like, As soon as you start questioning the actual user experience of what a video producer and motion designer has to do, you go like, oh my God, that's horrible. There's so much potential to make this better. Anyway, I'm getting on a thing, but prior, was great. So that was the first one. And what was the reception of that like?

It was great. It was actually the first tool that I had made actually. And then I just went through multiple iterations and then I got the feedback from the team at Verasity as they were using it. So I rebuilt it from scratch and then I learned how to an extension that would work in both Premiere and After Effects. So that was new. And then once it was polished and had a nice UI, was like, you know what, let's see if the wider audience is interested in this. So put together some promo materials, submitted it. Yeah, and Lloyd loved it and he was really excited to help hosted on the platform. we got that listed on a scripts and yeah, it's, it's, ⁓ you know, it's, it's had a really decent reception over the years and stuff. And like lots of people are still using it ⁓ and stuff. So, so yeah, that was our first, my first tool that I released publicly. And then that opened a bunch of doors, got to meet more people ⁓ like Zach love it. ⁓ I met at SIGGRAPH and we were chatting from that and he, got me plugged into the whole Adobe dev community, which is fantastic. He's done.

Yeah.

a lot of work around documenting all the undocumented stuff for Adobe apps. yeah. ⁓

So have you now. We'll get into that. Yeah. Amazing. Um, yeah, I mean, it, it seems like a perfect situation you're in where you're both doing the custom stuff with a team. You've got the, like the, frameworks and you've also got products as well. Cause just like you said, like you're working on the stuff behind the scenes and then you're like, okay, this is like somewhat validated, but let me just check whether the market wants it. And they're like, yeah, they do. Did you know, uh, Lloyd back then? Like, or did you in, can prove that was the first time.

Mm-hmm. Yeah. No, that was the first time. Yeah, first time meeting them and stuff. yeah, we've worked on several projects since we built the UXP framework for them. I think it was two years ago. ⁓ So now if you build a UXP plugin on a script, you're using Bolt UXP with a scripts licensing framework and everything wrapped around it essentially. ⁓ but yeah, that was the first time meeting them.

Amazing. Amazing. And, and back then as well, like, so you mentioned, ⁓ Zach Lover, but you said that, you know, he did a lot of work to make the space, ⁓ more documented. Was it really sort of messy and unknown? Did you have to have a certain amount of appetite to really get started? Like if you compare now, like the people who go to your, who in your discord or like, who know about developing, ⁓ promotion and like all of the, obviously the language model stuff that now has access to all this.

Mm-hmm. Yeah. Yeah. Mm-hmm.

Like when you compare the difference, like it seems to me you would have to have had a huge appetite to be able to do that back then. And it would have taken a lot of like roots just to get like a small little answer. Like what's the this for that. So I can carry on writing this thing. Maybe you could just talk to that a bit.

Yeah. yeah, absolutely. Yeah, there was, it was the Wild West for sure. And I mean, it still is in some sense, in some ways, but with the docs for Adobe and the types for Adobe, those two projects, it's just so much better now than it was. ⁓ But yeah, in the early days, was a lot of hunting down old forum threads from years ago and emailing Adobe reps directly and ⁓ just asking and then trying to find where these communities were. And that was a big thing that helped me because again, I was the only developer in, company at the time or was doing anything like this so didn't have other people to bounce ideas off off of. ⁓ So getting plugged into the Adobe dev community was fantastic so yeah huge thanks to Zach if you don't know Zach he's a great guy who's you know really helped the motion design community and the dev community ⁓ in tooling and stuff in so many ways. ⁓ But yeah getting plugged in and then having like a place you know you can ask questions and then eventually start sharing resources and stuff and ⁓ that was that was fantastic so definitely And it's nice, yeah. Now today we have the Discord, and so that's what I'm trying to make as well for people getting started today. So that's a lot easier path than... I think we're just over 600 people in the Discord. Yeah.

Is that- what's that got up to nowadays? sorry. No, what's Zach up to nowadays? think we've got lost.

what is that? What does Zach have to do? Yeah, I'm not sure. I haven't chatted with him a little bit, ⁓ but I know, I think he's still doing a lot of, you know, custom tooling. I think he focuses more on the graphics packages and toolkits. So like a lot of motion graphics template automation and that sort of thing. At least that's what I see from his stuff. ⁓ But yeah, he's done all sorts of stuff. Smart guy. yeah.

Nice. All right. Threading back to where we were just in your career. Like, have you worked at Buck? I got that wrong. Or you're a creative technologist, right? Or what was the title?

Yes, yeah. Mm-hmm. Yeah, I was a creative technologist. Spent most of my time on pipeline development, but my title was creative technologist. And so that was actually, yeah, that was a connection that Zach got me in touch with them because they were looking to expand their team. And yeah, after doing ProIO, that was my portfolio piece and that was my first kind of entry way into all this stuff. And so, yeah.

Mm-hmm. Mm-hmm. So did people at Buck use that? Was that sort of part of the thing? It was like, this is the guy who did pro IO or do they have their own like more sophisticated kind of pipeline management tools?

⁓ Yeah, they did. think a few of them had used it before and were aware of it for sure. But that was definitely a piece of like, yeah, you're on a scripts and you have these good reviews and everything like, okay, this is, ⁓ you know, this is something that we could, you know, work with and everything. So that was, that was cool. I was able to, and then of course, like, it's, there's two parts to this job, right? One is not just the development, but also knowing the workflows intimately and knowing how.

Great.

you motion designers work and how video editing works and the best practices and all that kind of thing. ⁓ It's kind of, you know, the, where those two fields come together is, you know, what really works best for this type of tooling. ⁓ Cause you can build features all day, but if they're not useful things for artists, then it's not, it's going to be off, you know, missing the mark and stuff. ⁓ So, so yeah, so that's, that's how I got plugged into Buck. work there for a while. Fantastic experience. A lot of great minds. ⁓ Michael Dilling,

Yeah.

and a bunch of others. ⁓ Yeah, really smart people. So I got to build for a much larger team. So we got to build, you know, all sorts of tools, some just for like one-off projects, ⁓ some more like larger baseline pipeline tools for the whole team. ⁓ But yeah, was able to, yeah, learn quite a bit there on, you know, how to build projects on a much larger scale, which is fantastic. Building for more apps outside of Adobe, know, Cinema 4D, Figma,

Mm-hmm.

⁓ even some Nuke and Maya and Houdini and all these kinds. So yeah, it was a a great experience.

Nice. Yeah. Maybe we could just like step back and hang out here a little bit, because I think a lot of people who I know, and certainly myself, don't appreciate like what pipelines are in larger companies. You know, people who just deliver one or two videos here or there, or a few versions or things like that, or maybe have two to three people in their team, but maybe it just gets to a totally different scale. Maybe you can just, like the word pipeline management to someone who hasn't worked in a large agency like Buck probably seems like overkill, but it's not. And I would love to learn, sort of get an education from you on what that kind of entails, how big it can get.

⁓ Yeah. It's funny. It's kind of a nebulous question because the word pipeline means so many different things to different people. So you say, you build a pipeline? It's pretty much meaningless. It could mean a hundred different things. But I guess generally what it looks like. If you're a smaller company, a lot of times you'll have one-off tools for small little pieces of your workflow maybe. ⁓ Generally when you're for the pipeline, it's talking about something a little bit bigger, a little bit more, you know, covering a little more of the workspace than just like, ⁓ or sorry, the workflow than just saying, rename these layers or export this one thing or do this one importing step or whatever. It's usually gonna oversee like a larger part of the workflow. And for some companies, ⁓ they have pipelines where you don't even touch the file system, you do everything through their pipeline tooling. And some it's like, okay, well, we have this special need for when we're trying to get stuff from ⁓ animate into After Effects and we're trying to do stuff from Photoshop to the web or whatever. And so they have this whole kind of tool set that fills in all these different gaps. ⁓ So at Buck anyway, it was a combination of both. We would have one-off projects where they would say like, we have this project for Netflix we have to do and we need to get this stuff from. these like strokes from ⁓ animate into After Effects, we need a special tool for that. And then other times it'd be like, okay, let's restructure our underlying pipeline tool that's gonna kind of manage how we create and version projects and publish and all this kind of stuff. yeah.

And can you like, it seems to me like just in my very small world that there's always a tension between like, do I just do this now or I do it a few times for this project or do I step away and try and automate this? like, I just work that out by gut feeling. Is there a more like process driven, like quantitative way of like doing that? Because you can make anything better probably in a team, but like, how does that get queued up? Sometimes people unsure like, yeah, maybe you can speak to that.

Yeah, it's like the meme that says, ⁓ why spend 10 minutes doing something when you can spend five hours automating it, right? But yeah, I like to say, if it's something that you are doing consistently, then that's up. exactly.

Or in my case, sometimes failing to automate it and then going back to doing the thing again.

Yeah. So yeah, that's a good meme. But in seriousness, definitely if it's something that's reoccurring, or if it's something, even if it's not reoccurring, it's something that's gonna take up a lot of time. For instance, in some cases it was like, hey, there is, we have this project that, yes, it's only gonna last for a couple months, but the amount of time it's gonna take our artists to do this manual step of the process is going to just take like weeks and weeks, basically. So spending one week building a tool to automate it is going to like, you know, save us all this time and money basically on the long run. So that's one way to quantify it. If it's like a project-based and the other thing is just of course if it's ongoing. Hey, we're always creating the same, you know, projects from the same folder structure. We're always publishing to Frame.io or to these Google Drive folders or whatever. Or we're always, you know, creating templates from this CSV spreadsheet or this collective document or whatever, your data structure looks like. Anywhere where there's repetition like that, there's room for automation. So yeah, it really depends on the company. Yeah, go ahead.

And it seems to me like, sorry. Yeah, I mean, so my, company I'm at at the moment, Glide, like the whole idea is like you build internal tools ⁓ yourself. ⁓ It's like a, ⁓ platform that lets you build ⁓ primarily on top of spreadsheets, but any kind of data source. And what we inevitably like see in, in even really large organizations is that you kind of got this rolling set of different processes, obviously inside of a department or an area that are kind of those things where like it was set up by someone else a while ago. And that tool is there and everyone has to go there. And because again, like you're doing a thing, you're not going to go and like question the things, especially if you're not technically oriented. And so I imagine that like, if, if like internal companies have the resources, they have set up all this stuff. I'm rambling here. I guess I'm trying to say is like, when you come into a company, do you often find that there's like a load of these kind of like skeleton weird ways of working and like you're sort of part of what your job is like pulling it together or like remaking a part of it.

Mm-hmm. Yeah.

Or is it usually like, needed this to do this. Like you coming in and cleaning stuff up or are you like making a thing.

Yeah, we get it all. Sometimes people come in and they're like, they know exactly what they want, which is nice. Like, hey, we need a plugin that does A, B, and C. And we're like, cool, we'll put together a quote for you, easy. Sometimes they're like, we have this problem. And we're like, okay, let's jump on a call and talk about the problem, see how we can fix it and the different ways. And a lot of times that brings the servers like, you didn't know there was this feature in After Effects and this feature in Premiere that would wake this part of your workflow better and stuff. ⁓ And yeah, sometimes it is rebuilding tools. Hey, we had a guy build this five years ago, but he's not yet the company anymore and we don't know how to maintain this. Can you guys rebuild it or whatever? yeah, we get all sorts of requests for sure from really specific to really general and stuff.

And presumably you have lot of repeat stuff, right? Because once people get into the world of like understanding that they can do stuff, all of the ideas pop in and they go, I want more.

Yes, absolutely. Once that kind of switch clicks in your brain and you realize what's possible with automation, then there's like, well, we could be automating this. like, OK, we don't want to pay people to do this task. Let's have this part be automated so they can focus on the stuff we actually want them to focus on and stuff. ⁓ yeah, repeat clients are really common for sure once it's the right people in the company understand the power of automation, what that can really save. yeah.

Mm-hmm. Mm-hmm. Mm-hmm. Mm-hmm. Mm-hmm. Mm-hmm. One final thing before we kind of leave back in that kind of world of large agencies and internal tools and stuff is, so would you, what's a technical director in your mind? Is that different to someone who does what you do, whether you were inside a company or outside of a company?

Mm-hmm. Yeah, again, these terms are all kind of loose. It's funny. I've talked to a lot of people like what's your job title? What's your preferred job title? Yeah, technical director is one, creative technologist is another one.

Sure. Yeah, yeah.

creative tooling or, you know, two, I think I've heard like 2D ⁓ technical director, 3D technical director. Yeah, there's a bunch of terms, I guess, and a lot of them mean different things. ⁓ But yeah, think generally it's ⁓ any term that has technical in it generally means, you you're either using some sort of development, whether you're coding it yourself or you're, you know, using more technical node-based tools or something to achieve different automations. And that's kind of your mindset. mindset. Whether that's your full-time job or half of your job or just a fraction, ⁓ you could be a technical artist, technical director. ⁓ Again, it really depends. If you're looking at jobs and you see the technical director, just read what they're actually looking for and that'll tell you what that term means to them. It's pretty

So yeah, I'm interested then about the tipping point about when you went full time with HyperBerry and presumably it was you just yourself to start off with. So yeah, what was it like that first year, just you if it was, and how did you decide to like go full time?

Yeah, so a great question. ⁓ it's, you know, either working freelance or starting my own company was always something I kind of had in the back of my mind. You know, it was super fun to work on, you know, work at a big company like Buck, work on projects for large clients and, and have fun with all those challenges that come with that. But eventually, like I wanted to have something that, you know, the work I did day to day was something I could, I guess, take with me and like be working towards a goal and be building something. So I was so yeah, Hyperbrew was a side project for years. ⁓ But then once it was feasible to go full time, ⁓ know, put in my notice there and just went full time on Hyperbrew. And yeah, it's been it's been fantastic. And anyone who's gone freelance or started their own business knows there's a lot of ups and downs, you know, because you're having you're not only doing the work, you're having to get the work. And so you're wearing all the hats being a salesman as well as, you know, being an artist or technical director or whatever. it is you're doing. ⁓ But yeah, all of the rewards are well worth it. So yeah, it's, ⁓ but yeah, started out with just myself and then, you know, started slowly adding more people to the team to help with development, with UI design, now with marketing, which has been immensely helpful, and all sorts of aspects of the company. So yeah.

And you had a shift, I think, I can't remember where I read or heard about this. Maybe you told me about it, about shifting your kind of messaging. Cause like as part of your marketing, there's a shift in messaging and it's kind of gotten a lot more direct or like clearer or more effective recently. Maybe you could talk a little bit about that. Cause this stuff is like nuanced. see it in like custom internal tools for businesses is what Glide sort of focuses on. You know, like at the moment we've got a big brand campaign coming up, which is called start with a spreadsheet. And it's like all about like start here at the most simple,

Yeah. Totally. Sure, yeah. ⁓

primitive of your business and then get into the complex stuff. I feel like it's probably similar for you some ways.

Mm-hmm. Mm-hmm. Totally, totally. mean, you have to know how to do the thing, the service or the product or whatever it is that you're offering before you can get into the business and marketing side of things so you can deliver. But then also, you have to know where the next job's coming from and you have to start putting your company out there and communicating with people in ways that make sense to them. So yeah, so for instance, the first tagline HyperBrew had was ⁓ software solutions for video. Super vague, like what does it mean? Who knows? It was terrible. ⁓ And then over time, started to improve that a little more. Took some business courses and started learning from different books and resources online and stuff. And so the V2 was custom plugins for Adobe apps. So that was a little bit closer. So it describes a little more what we do. We do custom plugins and we do them for video. or for Adobe apps. then, yeah, and then we actually use that the first time we went to Adobe Max in Adobe Max 2024. And it's funny, the most common question we got is people walk up and they'd be like, custom plugins, what kind of plugins do you make? Or like... what are custom plugins? ⁓ And so then we have to get into describing like, here's the types of plugins we can build, here's the automate and everything. But it was kind of asking the wrong questions. It was like going the wrong direction. It was kind of getting hung up on like an aspect of like the how, not so much the why. What's the big why are we building these plugins and stuff? ⁓ So yeah, I brought in ⁓ a marketing expert, ⁓ Eric Moore from Brand Autopsy. out, he's an excellent mind for this stuff. So he helped kind of interview

Mm-hmm.

bunch of our clients look at the industry and the space and figure out what we should really be positioning our facilities as. So our updated tagline for this Adobe Max 2025. ⁓ is we automate the boring, you create the amazing, and then custom plugins for Adobe Creative Cloud. So that really kind of hones in on like, you know, the why, like sure you build plugins, anyone can build plugins or whatever, like what's the deal? What's why is important? Oh, we're automating the boring stuff so you can create the amazing stuff. So that's kind of the emphasis, like, you know, there's, that's why we do what we do, and that's the types of results you can see when you get custom automation built and stuff. So that really resonated with people, like it was kind of cool to see the difference

Mm-hmm. Mm-hmm.

between our booth at Adobe Max 2024 and then last year 2025. And ⁓ so people kind of came up asking the right questions a lot more like, can you automate this? Can you automate that? Like, I hate doing this in my workflow. Can you guys automate this? So that's kind of like more the direction and like how we do that doesn't really matter. Is it a plugin? Is it a desktop app? Is it a website? It doesn't really matter how we get there and we can build all sorts of stuff, but yeah.

Mm-hmm. Mm-hmm. Mm-hmm. It's so interesting how language is like, just, think about this all the time, language, you know, when you do a little bit of development or when you do a bit of anything, you go back to your normal life and you think, there's a bit of that and that. Like, so like a term can like program people like a H1, you know, like literally changes the behaviour and the mindset and the context of someone who comes up to your booth. And obviously you would have gotten a real visceral, not visceral, like, yeah, in-person sense of like,

Mm-hmm. Exactly. Yeah.

what that is with people and like in this digital world, it's so easy to not feel that. You know, you might look at your analytics, but you don't really get a vibe or feeling as about what your H1 is doing. So that sounds really good. Do you feel like it's there or do you feel like there's another tweak coming or like, I don't know.

Mm-hmm. Right. Yeah. Oh, I'm sure there'll be tweaks along the way. I'm not saying it's a, it's an ever evolving process for sure. But I know it's a lot better than what we had and stuff. And so who knows? We'll, we'll see. We're, signed up for Adobe max 2026. So we'll see if it's the same or we might tweak it. Who knows? We'll see. But yeah.

Yeah. And so, yeah, you're, you're a partner. mean, you're more than just a partner of Adobe. You do a lot for Adobe. Maybe we could talk a little bit about that without getting too much into the open source stuff. We could like touch on that sort of thing. is it, yeah. How are you involved in Adobe? I mean, there's Express. There's, there's obviously the way that your products like fit in their ecosystem. Maybe you can talk about the relationship before we get to the open source stuff.

Totally. Yeah. Yeah, totally. So after, excuse me. After building ProIO and doing some community contributions, Adobe offered HyperBrew to be a part of the ⁓ video partner program, the Adobe video partner program. so basically what that means at a baseline is just that Adobe trusts, like Adobe sees your company is adding value to them. So you're building tools or offering services that makes their products more valuable. So obviously that's one. And then two, that they trust you enough to then put your logo on their website and potentially send work your way. So that's kind of like the baseline of what a video partnership looks like with Adobe. ⁓ But we really take it to, we take it a step further as far as we can in a lot of respects. So it gives us better access to a lot of the teams at Adobe. So we have ⁓ several of the video and photo departments at Adobe that we connect with regularly and check in with and test different beta builds and have calls about new APIs and that sort of thing, especially with the migration from the the older CEP extensions to now the newer UXP plugins. There's been a lot of changes and a lot of like, hey, we're rebuilding how plugins work in these Adobe apps. What should we change? So we've really tried to maximize that. In some cases, it's been paid help and other times it's unpaid, but it all helps us in the end because it's making the tool set better for us to then use in the future, which is what we're, this is the stuff we program with when we're building client tools. ⁓ So if we can do some meetings here and there and provide some examples of how to improve things, it's going to just help us on the long run. yeah, it's kind of a full-time job staying in touch with all the different departments because Adobe is huge. there's like ⁓ so many different teams. There's Premiere, After Effects, the UXP team, Express, Photoshop, and yeah, list goes on. But well worth it for sure in a lot of ways.

Do you have a sense without saying anything of like any kind of future direction or is that kept very secret? Do you just understand what's in the betas?

Yeah, I think we'll see stuff that's ⁓ not quite in public beta a lot of times. So like a lot of times when they're testing, especially with the new APIs, they'll have it, you know, either with... Yeah, non public beta or it's like, you know hidden a hidden feature in the app that you can unlock if you have the right Access and stuff like that. So so we'll see we'll see, you know stuff like that or they'll be like, here's like a prototype You know of how we're gonna now do this thing in Premiere Can you guys test it out? Tell us what you think and so we'll test it out and then talk and say like, okay Like this is working but like this is confusing or I feel like trip people are gonna trip up on this You know, how could we improve this or ⁓ extend scripted it this way, but everything you speak and do it so much better you know, now's an opportunity to change. ⁓ So not as much like bigger picture vision, I guess. I'm sure that's probably, ⁓ you know, a little more secretive as to what's going on there, but more like immediate, like, okay, here's the new features we're looking at rolling out and here's the APIs around that stuff, so yeah.

So you get your own clients, obviously through your own marketing, through repeat people, Adobe send you people. I'm interested like what are some of the best, most interesting standout, most weird kind of custom tool projects without giving anything away about NDAs or anything like just interested in like maybe two, maybe three, like the kind of just give us a broad range of what custom tools are and what they can be.

Mm-hmm, yeah. Sure, So yeah, most of the time it's going to be plugins for the four big Adobe apps. That's Premiere Pro, After Effects, Photoshop, and Illustrator. And that's where we spend the majority of our time. But then we also get requests for apps outside of the other Adobe apps, other non-Adobe apps like Figma, DaVinci Resolve, and... ⁓ Yeah, I'm like, I send them a 4D. I mean, yeah, more for Adobe Express these days. But yeah, that's like the add-ons, plugins, extensions, that's kind of our focus. But I mean, we've done all sorts of stuff. We've done desktop apps, web apps. We recently got a request for an Apple TV app. So that was a first one. But a lot of times when you work with a client on a more typical project, and then they'll be like, wait, can you guys also do this? Or we have a need for this other type of weird automation. We have to get this thing from here to here. ⁓ So yeah, we work in all types of contexts. But as one example, there is a project we have on our website ⁓ that it was a looper tool for Premiere Pro. And so the client needed to film ⁓ like hundreds of their products at the start of every year for their big event that they needed to show off. Hey, here's all the new products that we have. So they had them all filmed on a spinning table and they would have to, you know. spend late nights and weekends editing all these videos in time for the big event. And so they came to us asking, hey, can this be automated in any part of this workflow? And so we took a look at it and we're like, yeah, we can use basically image recognition to automate the looping editing points in Premiere Pro. And then we could also use OCR text detection because they had a slate that said, it's this product and it's this color and stuff like that. So we could pull that data out. So basically, instead of having to have editors you know, do all the grunt work of like opening all the files and finding the edit points and typing in all the logos. They could just point the plug into a folder and it would go through all the videos, make all the perfect edits in Premiere, the perfect looping edits, ⁓ label them accordingly so it has the product name and everything, drop that into a motion graphics template, and then the editors could just jump in and start doing color correcting or fine-tuning or things like that. ⁓ And so that was one that was pretty, ⁓ yeah, definitely a weird one where like, this be automated? We're like, we'll find find out, you know. So did some research and we're like, yeah, okay, we can do this in these ways. ⁓ And then that saw massive savings. They said that was the first year the team didn't have to ⁓ stay late or work weekends for that project, because that's what they always had to do. So I'm like, if, yeah, so it saved a bunch of time and money and hey, made some people's lives a little better. So ⁓ that's always a good thing to hear, but yeah.

That would have saved a bunch. Right? Yeah. Amazing. What about the weirdest or anything or if there's not a weird one, like what is one that you've said no to or like you couldn't do?

⁓ Let's see here. I don't think we've said no to anything that we couldn't do like technically, unless it's outside of like our focus, I guess, generally. Our focus is automation. And so occasionally we'll get requests for things that are not really automation based or, you know, like, can you build a rendering engine for this? And it's like, that's not really what we do or that sort of thing. But generally it's, you know, if we're reducing clicks, if we're, you know, ⁓

Mm.

eliminating the tasks that a human would have to do. Anything that surrounds automation, that's kind of our focus and our specialty. So if someone comes to us with something that's totally outside of that realm, we'll generally refer them to a partner and stuff. yeah.

How low level do you go? Again, as a sort of semi-hoff, newbie developer, like I sort of lack the ability to clarify that question anymore. Maybe you can answer it well for us.

Mm-hmm. Sure. Yeah. Yeah, it really depends on the project most of the time automation in you know, the Adobe apps is JavaScript but occasionally Will need to use other languages sometimes C++ sometimes rest which are lower level languages And then in other apps, it's not JavaScript like in cinema 4d. It's Python, you know in Lightroom It's Lua and you know different apps to have different requirements and stuff. So JavaScript most of the time but

Interesting. Okay, cool. ⁓ Cool. I think maybe just stepping into the open source side of stuff now, like maybe you could just step back because you worked across a bunch of different tools over your time and some tools are more extendable than others. Obviously the Adobe ecosystem, particularly After Effects and Premiere are like a really extendable. And then I was thinking just like on the lead up to this chat video, I was thinking about other ones like Final Cut and it struck me that like there's obviously like hugely powerful product decisions for these big tools are like

Sure, yeah. Mm-hmm. Yeah.

Are we going to be a product first kind of tool like Final Cut or are we going to be like an ecosystem type tool? I guess like, do you see differences there? Like presumably like if, if Adobe wasn't as open as it is now, you would have a very, very different job or role that you wouldn't might not be doing this stuff. Yeah. Do you ever think about this and are there tools that you work with that are just like very closed? I, for example, DaVinci Resolve I haven't gotten into yet in terms of automation. Like, yeah, it'd be interesting to just step back and talk about that.

Yeah. Yeah. Mm-hmm. Yeah. Yeah, it's definitely a factor. How open, how extensible a product is is going to dictate a lot of times which companies adopt it and which... especially the larger companies. They're not going to want to adopt something that they can't plug into their pipeline or plug into their ecosystem and stuff. So there are some tools out there like ⁓ I know a big one these days is news and stuff is Affinity. It doesn't have API yet. So that's something that a lot of people want to build plugins for. think outside of like filter plugins, you can't do any sort of automation. That's one that's like, yeah, it's black box. Sorry, you can't do anything in there. There's other tools like DaVinci Resolve does have automation and we've done a few of Divinjee Resolve projects, but it's more limited than Adobe. So there are a lot of times we've had clients come and say, hey, can we automate this? And we're like, sorry, that part of the API is locked. Basically, we can't do anything there. If you want to do the workflow in Premiere, we can definitely do that same thing. And yeah, it really does depend on the app. And there is an API for Final Cut. We just haven't had requests for that come through as much. And I think it's newer and more limited. I think Avid as well has a newer API. ⁓ But ⁓ yeah, definitely if you have a product that opens the doors to whatever your company may need, ⁓ then I think it's gonna stick around and people are gonna say, maybe Premier doesn't do exactly what we need, but we can make it do what we need because it has the options for that. So yeah, definitely having a host tool that has extensibility, has an API is only gonna help for sure.

Mm-hmm. Mm-hmm. And so would it be fair to say that the largest kind of agencies or companies that are making video tend towards the Adobe ecosystem?

⁓ I mean the ones that we interact with, again that's because we how we position ourselves as well. So of course there's tons of companies who are using other ones.

Sure. Yeah. Could you make a bet on like whether that's true across the board? Like, I guess what I'm trying to ask is like, if you're a large team, a large pipeline, does that always lead to custom tools or, or like workflows or does it not sometimes? I don't know. Just interested.

Mm-hmm. Yeah, I mean, you'd be surprised. It's funny, sometimes it will get... all have conversations with managers of really large Fortune 500 companies and they don't even understand how automation really works or that it's possible to automate things in Adobe apps. And then other times there's small limbo companies that are like writing their own scripts and plugins and stuff. it really depends on the case by case. It's really based on knowledge. And that's a big part of our marketing is trying to educate people and say like, did you know that these tasks could be automated? You're not left with the tool set that you're given. You can build basically. We can build whatever you need to automate the tests that you don't want to be spending time doing. yeah, you would expect that, OK, if a company has this many people or more, they're having all of their boring stuff automated. But it's not always the case. It's not what we see. So yeah.

Mm-hmm. So I'm put you in educator role again for me and the audience as well, which is like, and maybe we could step back again, sort of to the beginning and up to now and where we're going with the whole like Adobe plugin landscape explained. think mainly for video. like we can stay in After Effects. You mentioned an interesting thing, what I didn't know about, which is like, there's another language, which is like Lua in Lightroom, which is really interesting to me. just assumed that that would be the same like sort of ecosystem, but not.

Yeah. Mm-hmm. Sure. Yeah. Right, yeah, you would assume. Yeah.

That's a separate tangent, like, yeah,

Yeah.

staying in an After Effects and, and Premiere world. So like people who've like watched, you know, the expression stuff from School of Motion. That was with Zach. I remember that was the first time I started doing any of this stuff. Like expressions is like the way that anyone goes into this world for the first time. think, okay, I want to automate something apart my workflow. Okay. There's this thing called expressions. Okay. There's a bit of code. I'll get into that. And then maybe you get into Extend Script and stuff like that. And then like.

Mm-hmm. Mm-hmm. Yeah.

You've mentioned CEP, UXP, like maybe you could kind of go back from the lowest level up to what UXP is and where are things going now and why there are all these acronyms flying around. I just make sense of it for people.

Sir. Sure, yeah, so at least in the Adobe ecosystem, there's always been a split, I guess if most of your viewers are After Effects users, ⁓ we can think of expressions as, the lowest level kind of form of coding, and not lowest level, but simplest form in After Effects, right? So expressions can, you can do a lot of stuff with expressions. It's always going to be tied to the timeline, right? So you're basically, an expression can only affect whatever's inside of its property. So if you add an expression to the position property, property. can do all sorts of math and linking and whatever effects you want on there. At the end of the day, all it can do is adjust what that property is going to be. So it's little code snippets, you know, that are limited to that one property. ⁓ Now, take it a level up. We have script UI or Extend Script scripting in After Effects, where you can do like simple scripts in a single file. They can have a user interface. They don't have to. They can also just be headless, which means they're just running in the background, essentially. But what scripts UI or ExtendScript can do in After Effects is automate different tasks. So now you can create layers, can create shapes, you can create comps, rename things, import, export, ⁓ all of these sort of automation tasks now become possible. You can even apply expressions or update expressions and that sort of thing. So it's kind of like a level ⁓ higher, right? ⁓ And that's also still, it's ExtendScript, which is a form of JavaScript, a very old form of JavaScript. Now, if you take it a step further, we have CEP extensions, which has been ⁓ which has been the industry standard and still is and on a lot of the apps still is for After Effects. ⁓ And so CEP extensions are basically script UI apps, but they have a web browser front end instead of a very old version of JavaScript for the front end. So it's basically Chromium on the front end and with Node.js, which gives you a bunch of capabilities for reading, writing files, making network requests and a bunch of other stuff, adding libraries from NPM. ⁓ And then on the back end is still ExtendScript, but now, so now you can build a, yeah. Sure, yeah.

So was pro IO just before we go on. Can I jump in? Was pro IO DQP or?

Yes, ProRio is CEP. So After Effects has the ability to do script UI, which is the more simple form like I described. Premiere doesn't, interestingly. ⁓ You can only do CEP. So because I started ProRio and Premiere, ⁓ CEP was the only option. So that's what I built it. then also CEP is a little bit easier when you want to add multiple apps into like have one plugin work in multiple apps as well. So yeah, so CEP is still makes up most of what we build today and stuff. ⁓ But that's slowly switching. And so CEP is being replaced with UXP plugins. So the whole kind of scripts extension plugin paradigm kind of falls apart a little bit with the introduction of UXP unfortunately. But UXP, yeah.

But am I right that it's not live yet or like that? I get confused. I was doing some research. Not in After Effects. I see.

Not in After Effects. Yeah, yeah, it depends. Each app is different. It's been in Photoshop the longest. ⁓ I mean, XD longer than that, but I think XD is discontinued. But yeah, Photoshop, ⁓ it's in Premiere. It's just now out of beta in Premiere recently. So a lot of people are starting to switch their tools to USP. ⁓ And then After Effects is not here yet, and Illustrator, it's not here yet. And then the other ones, I think, we'll see what happens with those. ⁓ But yeah, so UXP will be in After Effects, just not there. So it's something that people are kind of like planning for. If they only build for After Effects, they're keeping an eye on what's happening in Premiere and Photoshop. But yeah, but if you're building a, you you wanna build a plugin with a nice UI in After Effects today, CEP is the way to do it, so.

Why UXP? is, like, what's the TLDR there about, like, why that happened?

Yeah, good question. ⁓ There's a lot of reasons. think from Adobe's side, ⁓ Chromium was becoming too much of a burden to keep up to date and there may have been some licensing issues around there, I'm not sure. But basically having that whole part of the stack being required, ⁓ they wanted to do away with that and just build their own basically front end and system from the ground up. So what UXP is trying to do is basically make it, instead of having to have two layers, the Extend script on the backend and the Chromium CEP on the front end ⁓ is just having one layer. So you can do user interface and everything and all the backend calls all in one space, which does have its benefits. Unfortunately, there are a lot of drawbacks. Anyone who's built a UXP panel knows that ⁓ the UI, they can't do all the same stuff they could do in CEP. And we have ways around that with WebView and BOLT UXP to kind of... ⁓ know, bring back some of the functionality if people want to take that step. But out of the box UXP, ⁓ it's more simple UI and that sort of thing. So, yeah.

so then the final layer, if there is a final layer is frameworks and then bolts. So guess that leads us nicely into how that happened and where it sits now. What was the inception of it? What was the original idea? How has it evolved since then?

Yeah. Yeah, so yeah, Bolt CEP was our first open source project. And that was really, you know, something that I'd gone through so many iterations through over time. My first build system was... ⁓

So you've been building it for years.

Yeah, yeah. I mean, I've had different versions of it, I guess, for quite a while, but it would change forms, right? So the early days I was using Gulp, which is a very kind of like, build your own build system from scratch way of doing things. So you'd say like, okay, I'm gonna write a command to read these files and write a command to strip out these comments or add this formatting or whatever, and then output it to this zip. Every single step of the process you have to specify. they give you the tools, but you build the whole build system. yourself. And then I went through a few iterations with rollup, with also remember some of the other ones, Parcel was one version. ⁓ And then eventually when VEEP came out, that was kind of the one that really stuck. And I was like, this is super fast. The plugin system is super well-documented and flexible. ⁓ This is the one I want to go with. started building a really robust system just internally. And so we use that on our own tools. And a few people, few developers, you the community would ask, what's your workflow? I'd sometimes I'd share with one or two people, but it wasn't public at that point. But yeah, but you know, I really wanted there to be some sort of open source contribution for HyperBrew. So ⁓ eventually was able to, you know, polish that up to a point where I could open open source it and make it public to the world. ⁓ And that's been really well received by the community. We've seen a ton of, you know, developers, solo developers to Fortune 500 companies implementing it in all sorts of ways. We have, you know, Battleaxe tools, Plug and Play tools, now Good Boy Ninja tools, all built on BoltCEP, which has been awesome to see. And with open source, ⁓ now it's, the code is public, right? Anyone can use it, download it, it for free, and then they can contribute to it, which is well. So we have had some really good community suggestions and some PRs, which PRs are like, if you're not familiar, a way for you, it's a pull request, but it's basically a way for you to submit code changes to a code repository and say like, hey, you I think this feature could be improved this way and they'll do the work and submit it. And then you can kind of go back and forth with revisions and then eventually merge it in. And now the project is bigger and better and now everyone gets to benefit from those features. So yeah, Bolt CEP has really awesome for HyperBrew and it's also led to new work opportunities as well ⁓ in a lot of ways. So yeah.

Nice. Can you explain it? Cause I think some people in will understand most of that. And some people will like not know what Vite is or Vite or how you pronounce it. And maybe you can step back and explain like, what is developer experience and what are you trying to solve here? Like again, just talking back to the early days, like how did you, what was it like when you made Pro IO or even the first thing that you did or when Adam first made

Sure. Yeah. Mm-hmm. Sure.

a butt capper or whatever, like what was it like back then and how is it different now in like terms that a normal motion designer would understand. Cause I think this has, I always think about the parallel between like the developer experience for video, as well as the developer experience for developers. There's a lot of similarities of managing files and like dependencies and like tooling and yeah, so it's really interesting.

Yeah. Absolutely. Yeah, so the whole idea. Mm-hmm. Right. Absolutely. Yeah, you know actually that brings a great example. So if you're working in video, you know in After Effects or Premiere, you're working in After Effects Project or Premiere Pro project file and you have all of your dozens of files sitting all over the place that you're linking. You're not gonna send your Premiere Pro project file to the client if you want them to review and edit. That wouldn't make any sense. You're gonna export a single file, upload it, and then they can view it, right? So it's kind of the same idea. ⁓ In a sense when it comes to development is that you develop in one way and it makes a lot of sense to develop in ways where you can have, know, split up your code into different portions. So you're like, I'm gonna have a code, a file that's doing these functions and a file that's doing these functions over here. And then that way, you you can structure your code, especially when it gets big, like you've, know, tens of thousands of lines of code, you don't want it all in one file, it's gonna be a nightmare to manage. But when you go to ship it, you want it, a lot of times it needs to end up in one file or a few files. And you'd want to strip out a lot of the code that you're not using, if you're using libraries, you could be using one function out of a ginormous library and you don't want to include all that code you're not using because that's just going to slow everything down. So that's kind of what build systems were the early day ones like Grunt or Gulp and then now we call them like yeah packaging, package managers. ⁓ Bundlers is the term for VEET or Rollup or any of these kind of tools that basically are like, instead of you having to build your own build system from scratch, this bundler is gonna basically work out of the box and then you can tweak stuff as needed. So VEET, yeah.

Is bundling the same as, like, compiling, or is that a different thing?

Compiling can be part of the process of bundling. Really technically speaking, compiling usually refers to low end code being compiled to binary ⁓ machine code essentially. So you're writing in C++ and you're building to an EXE or a ⁓ program file, .app file for Mac. And then transpiling is the more technically accurate, even though it's less understood and stuff for JavaScript where it's more like. The input is JavaScript, the output is also JavaScript, but it's just a more minified, faster, concise version of what you're doing. ⁓ But same concepts where you're like, okay, I want to write the code in this one way, but then it needs to be outputted in a different way that works better for everybody.

suppose another analogy that might be relevant here is like building, because most motion designers have built their own rigs or mogets or motion graphics templates, right? Like maybe there's a similarity there between this or not, I don't know.

Mm-hmm. Yeah. Yeah, I mean there could be right in the sense where yeah an After Effects right you're building an After Effects project ⁓ and then you only want to make a few properties editable for like your client to use and say like a lower third or a know on-screen pop-up graphic or whatever so you can package that into a mo-grid and then they use on Premiere and now they can only edit the text or the colors or whatever they're not gonna accidentally move something off screen or whatever and break the whole template and stuff. So yeah, that can definitely be a way of thinking about it for sure.

And so getting back to Bolt, it touches a definitely, there's a number of different Bolts, right? Bolt Express, Bolt, and you've worked with Figma and maybe you could give us the landscape of what that evolved into, because there's a few different areas now.

Yeah. Yeah, absolutely. So yeah, Bolt CEP was the first open source project that HyperBrew released. ⁓ And yeah, really well received. And with the migration on Adobe's side from CEP to UXP, UXP was definitely our number two framework that we wanted to build. And at the time, a lot of the UXP stuff was very new, but we were kind of starting to keep an eye on how it was evolving, what it was going to look like. So that was one that we had on our mind. ⁓ But. Of course, allocating resources towards open source is always a challenge because it's not paid work, almost as I am. But that one was, we were able to fast track that project thanks to a scripts because they wanted a new licensing system to work for UXP plugging in Photoshop. So we're like, hey, ⁓ if you guys want us to build this, ⁓ we can basically build part of this as an open source tool. We'll list you guys as a founding backer of the project, and then we'll build your licensing framework basically on top of that. ⁓ So that was big things to Ascripts for making that possible. So we were able to jump start. making Bolt UXP. And so that's now released. So that works in Photoshop, InDesign, and Premiere Pro. And it will work in the other Adobe apps once those are released as well. ⁓ So yeah, and then after that, we were able to do, I think Bolt Figma was the next one. And so that was when Figma has a creator's fund, it's called, where you can submit ⁓ concepts for any sort of project that the community can benefit from, whether it's code or even like templates or whatever. we do get requests occasionally for Figma plugins. And that's something, I had it on the back of my mind. was like, you if we could build a better system for scaffolding a new project and then the interactions between the front and backend layers and then packaging and shipping and all that stuff. submitted, you know, said, hey, we built BoltUXP and BoltCEP. We would love to build a BoltFigma. Here's what we think it could do. And they were super excited about it. So they funded that project, which was awesome. So we were able to fast track BoltFigma, get that out the door. That was showcased at Figma's big conference that you every year, config. So that was super cool. And it's being used in

What year was that?

I believe that was 2024. Yeah, I'm pretty sure if I remember correctly.

Don't worry. Okay.

It might have been 23. No, I think it's 24. So yeah, so that was super cool to get that one published and then get, use it, other people are using it. And we've had some good community suggestions and contributions from that as well. And then Bolt Express was the last one that we did most recently. And that was through Adobe's fund for the Adobe fund for design. And so they, when they were launching Express, that's where a lot of their marketing funds were. And they were able to fund a lot of projects. developers who building add-ons to make the product more valuable, ⁓ know, increase the interest in all that. So I suggested, hey, you know, if we built a Bolt Express, well, how would, you know, would you guys be interested in that? Basically just simplifying, streamlining, making the whole developer experience much smoother and faster for ⁓ Express add-on developers. So they were super interested and they backed that project as well. ⁓ So yeah, so open source, you know, it's cool because it's something you can build and then everyone can benefit off of it. and it does open the doors to a lot of opportunities you wouldn't otherwise expect. ⁓

Yeah, that's amazing. And do think if you were just ⁓ like a ⁓ tool developer for the marketplace, right? Like you weren't doing custom development, would it still make sense to do open source?

Yeah, absolutely. think it's a way for people to, I mean there's a lot of benefits to it. ⁓ You know, sure. Yeah.

Yeah, I guess that's what I'm interested. I don't see inside of that yet. Like from the outside, looks like maybe it's more beneficial to develop like a pipeline of like customers and people who you're collaborating with. But I don't see all of the other benefits because I guess it just seems like it's such a generous thing to do. Right. And you have like so many like people benefiting from it, but I'm trying to get a perspective of what it's like from your side as the maintainer as well.

Sure. Yeah. Mm-hmm. Sure. Yeah. Yeah, no, it's definitely work. mean, once you put your code publicly, now it's open to everyone else to say, oh, why are you doing this? Or should we do this better and stuff? So you definitely want to dot your i's. your teas and everything before you publish something on GitHub. But with that comes, you know, now instead of it's just you or your team, you know, looking at something, now you have suggestions from all sorts of people, all sorts of developers with different, hey, you know, I use Boltcp in this context or with this language and something broke. Okay, cool. Well, now we can fix that feature and make it better for the next time we build one or the next client project that we have. So, yes, you're giving something away for free, but then also, you know, you're You're now getting the help from the community as well to make some product. That's not just yours, but it's now kind of everybody's Bigger and better over time and you have you know, lot of a lot more brilliant minds working on it and looking at it So it's been so that's like one aspect is just like okay now you can make the project better than it would have been if you're just doing it yourself And then another thing is, you know, lot of it does open the doors to a lot of work opportunities. Like I described with the funding from Ascript and Adobe and Figma, but also a lot of companies will see it and they'll be like, hey, we started building a plugin in Boltcp. We like it, but we really want our developers to focus on the main product. Can we just hire you guys to build the Adobe plugin for us? We're like, absolutely. That's our line of work. So it does open the door for a lot of those types of opportunities as well. So yeah, even though it is a lot of work to maintain and release and all that stuff,

Cool. So moving towards the end of this, I want to ask obviously about the future of Hyperbrew, where you're going, the way that the whole AI kind of development landscape is fitting in. But maybe we start in a more grounded way first, which is that you mentioned that you've done some plugins, custom tools for people which have involved things like OCR and ⁓ image recognition and stuff like that. Are there any other aspects of the custom tools or things that you see people using that use AI in a way that's like quite interesting, even language models? We can get onto

Mm-hmm. Mm-hmm.

klutz.jpt in a second, but like, yeah, are there any custom tools

Totally. Yeah.

that use like interesting AI that's beyond what people think of like as just standard LLMs?

Um, yeah, so I mean, OCR is a great example or image recognition. These are, you know, forms of AI that we've been using for years. They just didn't have the label on them. So we weren't thinking of them in that context so much like rotoscoping or other things like that. Um, but yeah, we, there's definitely, yeah, there's definitely a lot of cases where it's like, oh, some sort of tool that uses a form of AI makes a lot of sense to solve this type of problem. Um, especially, you know, where the variables are just kind of unknown, um, what the input's going to be. Um, a lot of times some AI is good detecting patterns and identifying stuff and recognizing stuff in that context. ⁓ One request that we did have come through was for basically a custom captions generator that was kind of fine tuned for a specific company. So that was interesting because they were using generating captions in Premiere, the default feature, but like lot of the name, like it was for a sports league. And so a lot of the team names and the player names, it would get wrong. the editor would have to go in and manually fix all these things. ⁓ course they have all this information they could say hey keep an eye out for you know these player names and these team names and all this and these like like sports terms and whatnot so they had a custom tuned ⁓ you know LLM that they built with you know their their own cloud service but they wanted that plugged into Premiere so we basically automated the same process you would do in Premiere with generating captions but it was via their customized there's custom trained ⁓ service for that and so that was one way where they could you know get a more personalized result that you couldn't get out of a typical off-the-shelf ⁓ render because it was trained on their data and stuff. So yeah, that does come up occasionally for sure.

So like you touch both worlds of, of, of custom tooling, custom development, and also off the shelf tools. And I just think it's a really interesting place to be because it seems to me it's like limitless what could be created as tooling for video, because the use cases for video or the things that, rather the, the, the things you want to put in a video and make video do are obviously endless. They're going to remain endless. And at some point, video will just become screen or what we see in our glasses or whatever it is in the future. So there's always going to be this next frontier of like the tooling and the tooling on top of the tooling and the workflows on top of the workflows. Um, I guess like my assumption is that you can't have one tool to rule them all. Right. Like you can't get because, and, and, and maybe inside of that question is like, do people want that? Like I look at big tools like motion for, I think it is at the moment by Matt McGruff is like absolutely amazing in some ways. And then also for some people, it's just like, I don't want to go anywhere near that as well.

Mm-hmm. Yeah.

because it's just like this, it's trying to solve everything and there's no disrespect. think what that Mount Megraff do is absolutely brilliant. I want to get them on at some point, but ⁓ yeah, maybe there's something here. Like maybe you can speak to that. is really just, are we moving towards this world where everyone's just making their own custom tools either themselves with their own like, you know, in 10 years time when AI is like so much better that anyone can, any motion designer can make their own tools as they go. Or is there actually like ways of solving.

Yeah.

Yeah, I don't know. It's a vague, rambly question. If you've got any thoughts on that, it'd be really interesting to hear.

Sure. Yeah, is the ultimate goal a bunch of little tools that do individual things or one massive tool that does everything? ⁓ Yeah, it's a great question. And I mean, I have seen a lot of products that advertise that they're like the one tool to rule them all, which is great and props to them. I think it can be hard for a lot of reasons. It can be like, if you have one tool that's trying to do too many things, generally each component isn't probably gonna be as good as a tool that's just dedicated to that, ⁓ or it could be in some cases.

Or is that just an illusion? yeah, like, yeah.

⁓ Another problem you're gonna run into is like marketing. So if you say, hey, we have a plugin that does everything and something, yeah.

What? You have to do marketing. Damn.

Yeah, if you say you have a plugin that does everything, the first question someone's gonna ask is like, cool, what does it do? So there comes problems with that as well. ⁓ But yeah, think in general, yeah, we're definitely, I think with AI and everything, it's definitely opening the doors to more people thinking in these terms where it was just never even something that was on their radar. And then they see, ⁓ Clutch GPT can write scripts for After Effects. Wait, you can write scripts? Like if that's their introduction, that's great. Now they're thinking on those terms. We've actually had a few jobs come through where someone was like, oh, we use AI to build this thing and it kind of works, but we don't know what we're doing. So can you actually build the real version for us? we're like, yeah, happy to help. But we definitely see more cases where that's led to at least opening, kind of unlocking that in people's brains and saying, yeah, this is a possibility. You can't automate this stuff. And yeah, I guess for us, mostly we're focusing on specific problems. That's most of the tools that we're building, whether it's our public stuff or our custom stuff. And again, custom tooling is our focus. So you're not going to see a bunch of products on our website. Most of it's going to be case studies and a lot of the stuff that we make specifically for teams. a lot of times, off-the-shelf tools are only going to get you 50 % of the way there. ⁓ But when you need 100 % of the process automated, well, that's when custom can really help, because it can be built exactly for what you need and stuff. So generally, it's for specific problems. That's where we spend our time. Yeah.

I guess I like, I, I'm always thinking about like, again, that tension that I mentioned about, like, I'm doing a thing here. I need a thing to do the thing. Let me go and develop the thing and how fast that gap can get for simpler workflows. And like Cluts GPT obviously was like a really first like shotgun fire. I just like playing around with that. It's almost a joke. Like I'd love you to hear you talk about like the origin of that, but like, yeah, when I, when I see like occasional Google demos or like, think it's Gemini or something that just like is able to like write.

Sure. Yeah. Yeah. Yeah.

web UI, like insanely quickly in response to a question, my brain, cause I'm a motion designer goes to like, well, that would be cool if you had like, you know, instead of having 17 panels infinitely docked in After Effects, is horrible. Why not have one window that can become anything? And like, my hope is in the future, that is the future of motion, but I don't know whether that brings up any thoughts. So that and Clutch GPT, I'd love to hear your thoughts on.

Mm-hmm. Right. Yeah, well Clutch GPT was definitely an experiment. was like, started building it as soon as OpenAI opened up their API so you could build apps with Chat GPT and everything. So that was like, hey, let's plug this into After Effects and see what happens. And no guarantees, it's labeled Clutch GPT, the tagline is the most unreliable way to write expressions and scripts in After Effects. And a lot of people didn't like that, they got offended. ⁓ we're being honest, you're gonna run into issues. We're not recommending this as a good way forward. But if you wanna play around and see if you get stuff that works. That's great. And Jake did a video on it that blew up and so that's our most popular tool today. It's free and it's had like 35,000 downloads I think at this point. but yeah, again, it's it has really opened the door to a lot of people thinking in these terms and they weren't, you know, they didn't even know this was a possibility before for sure.

Smile. Mm-hmm. so I guess another angle on this might be to talk about like the agent side of stuff. So one thing that happened last week, I think was a remotion with Johnny burger or Berger. Don't know how pronounce it. He's coming on actually in a couple of weeks time. ⁓ he opened up, like a Claude skill to allow anyone to like, or any agent really to make videos with Remotion. I don't know a lot about Remotion. That's kind of why I'm getting him on. It's been around for a while, but like that sort of blew them up really in a way. It's like allowed people to interact with it. guess thinking of, I don't know, Bolt is obviously a very different thing. Clutch GPT is obviously a very different thing. It's not that, but like how close, far away, what would it take to get to a point where we can like ask systems to do stuff in After Effects for us? ⁓ Like how, what... It feels to me that that's like tractable, but quite far away because there's a lot of machinery around that. Maybe you can talk to that a little bit.

Mm-hmm. I mean you can today. It's I wouldn't recommend it, but there yeah if you get if you get if you go look on YouTube there are yeah, there's there's plenty of tools and DIY plugging chat GPT directly into After Effects and telling it to do stuff and it does stuff ⁓ Yeah, it's it's unreliable. You know you can't guarantee what you're gonna get, but if you want to play around with that. That's definitely

Okay. Well, there you go. There's the point. I saw a demo the other day as well. was underwhelmed. And is that the fault of the interface into it and what After Effects can give out? Or is it a fault of the models or a bit of both? Is it just that we're not there yet in terms of like language models, understanding and visual recognition and its artisticness, you know?

Yeah, I mean, I'm not an AI expert myself, but I mean, I've definitely spent many hours and days and weeks into it and been testing before Chachi Pati was around and everything. ⁓ But there's, from my understanding, there's some underlying issues to how LLMs are constructed today that... they're just not gonna tell you whether they are confident or not. So that's a challenge. You can say, okay, generate a script for me and tell me what the percentage of the confidence is, but it's not actually gonna know. There's no way for it to know the way that it's built. So it's a lot of trial and error. So there's tutorials online to build your own apps, build your own plugins with ChatGBT, but a lot of it is like try, fail, try, fail, try, fail, works. and then that works today, but is it gonna work tomorrow or is it gonna work when you plug in a different type of file? You're not gonna know until you try and stuff. So yeah, we don't use AI for any of the code writing process in HyperBrew aside from some simple inline code completion just because we're building reliable tools for clients that'll work the same way every time. But who knows what will happen in the future, we'll see. But that's the current state of things as we've seen it evolve, but yeah.

Cool. Just tying it back to like video and Figma and things like that. Figma acquired Weavey, the tool recently, which is going to be Figma Weave. Like, do you have any take or thinking about this, like how Figma is going to enter more into the video space and AI video gen and how that's going to come into it? Cause yeah, again, I'm always interested in the user experience. I did a three and a half minute film with Weavey a while back and it was great. ⁓ was fun. I wouldn't say I would release it was like an internal funny project, but it was like, it was very like cinematic. And I really got into filmmaking in a way that I've never done before because of, yeah, not actually doing loads of videography. ⁓ and that's really exciting about the future, but the user experience like was awful, not because of any fault of Weebies, just because it's such a new paradigm, like nodes. mean, you should have seen the node, like the node.

Mm-hmm. Yeah.

trees and spaghetti I had to get this thing working and the amount of copying pasting and luckily they have like JSON. So when you copy something, you can paste another JSON snippet or whatever to get the same tree. Yeah, I just wonder whether you have any thoughts there about like where Figma sits with all of this, because you were involved in that organization.

Yeah, so I guess I haven't had a ton of experience as far as like the video side of Figma. I know they are definitely making more, you know, adding more features in that respect and leaning more into that, which I think is great, you know, making a tool because I mean, Figma is so powerful as it is, it's, you know, online collaboration. It can do so much with vectors and rasters and all this stuff. So yeah, we'll see what that looks like. And, you know, people are actually going to start switching from, you know, NLEs to Figma for editing videos in the future. Who knows? But I mean, I think that's great. Making a tool that's online collaboration with more features like that, that's only going to improve productivity and open the doors for new types of collaborations and stuff. So yeah, but I mean, there are more tools leaning towards this. We're the all-in-one tool for whatever it is, whether it's like Blender, for instance, is a big one, or even DaVinci Resolve. trying to do video editing, 3D compositing, sound design, coloring, everything all in one. That's definitely a trend too. I'm seeing more, it's company adoption is slower and that sort of stuff. A lot of the industry standards are still the industry standard at larger companies and that's where the jobs are. ⁓ But yeah, we'll see. A lot of people are definitely trying to expand their apps to being like, hey, we're not just the design app, we now do video. We're not just the 3D app, we do color, but yeah.

There's just going to be so many more surface areas as well. think as well, like there already are so many surface areas now as a video producer. And I just feel like people often in my team asked me like, why do you stick with After Effects when you do something else or whatever? like, if you're not in it, you know, you don't realize just how many other things you need to pull in and what you need to do to make it work. And it is the most sort of extendable, useful place to bring stuff back to often. I remember like on the last interview, the interview I did with Adam from Battleaxe, I think he said something either on the interview or afterwards. He's like, I wonder whether my purpose in life is to become the Like build the tool that like joins together all the other tools or the layer manager, something like that. And it's like so many moving things that move between stuff and something just needs to join it together. ⁓

Yeah, Overlord is the key. That tool, that is an amazing project. Adam has put so much blood, sweat, and tears into that and it's phenomenal what it can do. Yeah. Yeah. So now he, I mean, he's built several versions of it, but the latest version of Overlord uses ⁓ Bolt CP for, you know, After Effects, Bolt, USB for Photoshop and Bolt Figma for the Figma integration. So it's a perfect case study for us. And yeah, he's even, ⁓

Overlord is the key. Did he build it before Bolt or like, or?

paid to have a few features added to BoltCEP that he needed for his project. So we have Battleaxe listed as a feature backer on that project as well. ⁓ But yeah, definitely it's people haven't heard of Battleaxe or they haven't heard of Atom, you say Overlord and like, yeah, well of course I use that.

Nice. Cool. Well, thank you again. I have like one last question, which is really about like what's next for you for Hyperbrew. And maybe another way of answering that question is like even beyond that, like what's a thing you haven't tackled yet that you'd love to tackle or that you see in the next five, 10 years that you, like what would your dream project be? That would be another thing. So like what's practically next and what would be amazing if it was at some point next.

Yeah. Sure. Yeah. Yeah, it's, it's, you know, hard to predict what's going to happen in the future and where we're headed. But I would say definitely improving our marketing and reach. Like that's been a big thing. Like definitely starting HyperBrew like had the workflow knowledge, had the technical knowledge. And now it's just about making those connections for clients and stuff. ⁓ And that's, that's a big part of what we've, what we've been doing last year and we'll be doing more of this year, I think. So just trying to, you know, communicate what custom plugins are, what automation is and how it can help to people who don't know that it's possibility. That's just so many these conversations we've had like at Adobe Max, which is cool because you can have these conversations in real time with real people. It's just, you know, people not knowing that this is even an option or that, hey, you don't have to be stuck 50 % of your day doing a task that you hate. Like you can automate that so you can then focus on the creative stuff or the exciting stuff for either yourself or your team. ⁓ And so yeah, definitely, you know, lot of our efforts are going to be in, you know, making that more, you know, educating people on that, ⁓ making that as clear and stuff as we can. I mean, mainly to grow the company and just, you know, so people can see that that's an option and stuff. ⁓ And then as far as dream product or project, guess, hard to say, we definitely like right now the focus is on custom stuff for us. So it's really, you know, whatever clients come to us with whatever needs they have. But if we switch more to product-based stuff in the future. I do have a few pretty big ideas that I think would be really successful tools. I know I would have loved to have back when I was doing editing and motion design full-time. ⁓ So we'll see those come to life, hopefully in the next few years. But ⁓ as of today, we're really focusing on the custom stuff, ⁓ just solving people's workflow problems and making lives a little more enjoyable.

Justin, thanks very much. It's been really enlightening and yeah, I've learned a lot. So thanks so much for your time.

Thank you so much for having me on. It was pleasure.