The past, present and future of Product at Synthesia

Shaping cutting-edge technologies and solving real-time problems in innovative ways can be incredibly rewarding - and this is especially true when working on a completely new category, such as AI video.

I had an insightful conversation with Guillaume Boniface Chang , Head of Product at @Synthesia about how we do Product. Keep reading if you want to know more about:

• The current state of Product at Synthesia and it's future direction
• The ways the Product team works and the challenges they face
• How Product and R&D collaborate
• The thrill of building product within a Gen AI growth-stage scale-up that has found Product Market Fit.

Our product at Synthesia is called STUDIO. Can you give us an overview of the current state of STUDIO?

I like to think we’re still building the v1 of STUDIO. We’re creating a new product category and there are many obvious functionalities we need to add for the product to feel complete. What we do have going for us is a strong product market fit and passionate users who love what the tool can do for them already.?

Part of what I love about working in STUDIO is the sense of possibility. There are so many product opportunities we will get to explore and build in the next few years (video translation, collaboration, video analytics, interactive videos to name a few). The challenge for us is to be smart about what comes first and execute well.

Here’s how Synthesia STUDIO looks today:

Could you provide some insights into the philosophy that shapes our product management practices and how that philosophy translates into our day-to-day work?

We’re democratizing video creation and that’s a super exciting mission. It forces us to rethink a lot of what's taken for granted in traditional video edition tools which are built for an audience of specialists. That puts a big focus on UX and UI not unlike what you’d need to develop a B2C product. A big win for us was to decide early on to mimic a presentation software layout, creating familiarity and a sense of ease for a lot of people with no video creation experience.

So we want our product to be approachable but we also want to push the creative capabilities to the max. That creates a tension that we resolve by following the “approachable on the surface, powerful in depth” principle. We try not to expose too much in the first layer of UI but advanced features are there and can be found when a user starts peeling out the layers and looking for them.

Animation or sound track management are good examples of features where we’ve deepened the capabilities without sacrificing the initial simplicity of our UX.

What are the traits that make a PM successful at Synthesia?

At heart we want PMs on the team to be builders and doers. People that are passionate about the craft and love spending time in the code or the pixels to get to a magical experience. It’s a high bar because it requires having built a bunch of hard skills throughout a career: coding, designing, analyzing data, understanding business strategy...?

It’s all worth it in the end when it makes the PM that unique person in the organization that can thread the needle between all the perspectives required to build a great and successful product. That also means every day is different, and we’re not shy diving in to unblock the team by putting together a mock, pushing a pull request or diving into our database to chart the way.

We’re growing the team at the moment. Can you share how the team will develop over the coming year and the reasons behind us making the decision to bring on new hires?

We’re still small. Myself included, we are 3 PMs working with ~35 engineers and product designers and 40 researchers who cover topics like NLP, Cmputer Vision, and voice. We plan to keep a high ratio of engineers to PM. The first reason is that our conception of the PM role doesn’t cover project management. Our colleagues are senior, engineering tend to work fullstack, they own and drive their projects and don’t require hand holding. That leaves us the ability to focus on Product work and cover a larger surface area. In turns it allows us to be autonomous in our respective areas and think strategically, with coordination overhead kept to a minimum.

We’ll continue growing the team thoughtfully and keep pace with the engineering organization. I expect a year from now we’ll have doubled in size.

We’ve come far since we were a Product team of one as of one year ago. How has the relationship between product managers, product designers, and engineers evolved in that time? How do they collaborate nowadays?

I loved those early days! I remember doing a bit of everything: writing requirements, doing designs, prototyping, user research … It was thrilling and eye opening. Things go so fast when you’re owning a project end-to-end! There’s no coordination overhead and that makes all the difference.

We’re keeping that idea alive in a number of ways.?

• First, we are extremely light on process, as we trust our team to decide on the best way to approach a problem. For example, not every project requires user interviews, especially if you’re building something that has become fairly standard across products.?
• Second, we’ve built a culture that emphasizes ownership and robust debate. One person is responsible for driving things forward but on their way they will get regular feedback from their peers. Depending on the nature of the project, that person in the lead might be a PM, a designer or an engineer. Key parts of the STUDIO experience are developed by a designer, including the product requirements, because we need someone with a creative mindset to crack it.

Our 40-person R&D team helps build the proprietary technology that powers our product. Can you share how the Product & R&D departments collaborate together?

We work with Research in two ways:

1) On the one hand we try to imagine which capabilities could unlock new use cases for the product and on that basis try to shape with them research proposals.?

It’s fascinating and also super hard, both because research involves a lot of unknown, but also because the technical landscape is evolving super quickly. Our gesture feature was born from this push and pull between the desire to make the avatar more expressive, and the technical insights of the research team.

2) On the other hand, we come at the tail end of a research project, once they cracked the problem, to imagine the UX that’ll bring that new capability to users.?

We work through the detailed requirements for the tech to be ready for production and provide continuous feedback on research results to get it to 100% readiness.

How is the Product team leveraging generative AI to shape the direction we’re building the product in?

Setting aside the stuff we do in the video space, which is our bread and butter, there’s a huge opportunity opened up by the latest breed of LLMs. We envision a new modality for creative products allowing a user to work at a different level of abstraction. We call this workstream “Assisted Creation”.?

We’ve already built a functionality allowing a user to go from a short description of their topic to a full draft of the script and the visuals for the video.?

We want to take this much further and turn it into a creative assistant that’s supporting you throughout your creative experience, suggesting and being invocable to perform a scoped edit for you. It’s not a radically new vision but for the first time in tech, it’s actually feasible to build something that’s more than a gimmick. It’s a complete greenfield and will require a bunch of iterations before finding the right recipe. We mean to be at the forefront of that.

What are some of the challenges someone shaping the product at Synthesia will find?

Besides the “Assisted Creation” workstream I mentioned. There are two challenges that I find fascinating at Synthesia:

1) We’ve built a video creation product that flips the paradigm of the industry.?

We have no timeline, in fact the user narrative is the primary way to drive the creation. Creating a new UX paradigm is exciting yet mind bogglingly complex. With the script as the timeline, how do we give our user fine grained control over animations and visual effects while keeping things simple and intuitive? It’s tough but good, you know? I think of building a paradigm as the thing you become a PM or a Designer for.

2) We’re pushing the boundaries of what’s possible with our avatar tech, creating human representations that are ever more expressive and realistic. When those innovations come to product, they create interesting UX challenges:

What’s the right level of abstraction to control those avatars? What should be automated, what should be configurable by the user? For example, how should we approach enabling building dialogs in the product.?

Those questions are at the core of our value proposition and a big part of our value proposition.

________________________________________________________________

Hope you find this interview insightful and interesting. In case you have any questions, feel free to reach out to me here on LinkedIn!?

And if you’re a PM / designer looking for an exciting opportunity, don’t forget to check out our careers page - we’re always looking for amazing new colleagues to expand our team!?