Why I Built React Video Editor

I started building React Video Editor while working as the founding engineer of The Data Refinery.

The goal was to automate product release updates as video for social channels. Instead of making a new video by hand every time something shipped, we wanted to turn the update itself into something we could share.

That got hard quickly. The videos still needed someone to adjust the timing, captions, visuals, and layout before they were ready. We realised that before we could automate the process, we needed a video editor.

I went looking for one we could use, but I couldn't find anything that fit. So I built one and put it out in the open to see if other people had the same problem.

Turns out they did.

I did not have a grand plan for React Video Editor when I released it. It was an answer to a problem in front of me. Seeing other people pick it up changed the way I thought about it. This was not just a missing piece in our workflow. There were developers and teams trying to bring video into their own products, and they were running into the same wall.

Those are the people I think about when I work on RVE now. They are not necessarily trying to build a video editing company. They might be making a tool for marketing, education, commerce, or something I have not seen before. Video is one part of what makes their product useful, but it is a large part to take on alone.

The first version of an editor can be exciting. You can drag something onto a timeline, press play, and see the idea working. The reality arrives later, when people bring their own media, make projects you never expected, and depend on the result being the same when it is exported. That is the part that has kept my attention.

AI has made the starting point much more accessible. The latest models can generate edits, draft timelines, and speed up a lot of the code. I use those capabilities and I am optimistic about them. They open up ideas that would have taken much longer to explore a few years ago.

At the same time, keeping video inside a real product is still hard. A generated timeline still has to play properly. What someone sees while editing still has to match the finished video. A browser still has to cope with the media a customer uploads on an ordinary day. None of this makes AI less useful. It just means that generating a beginning and caring for the thing over time are different jobs.

That distinction has shaped what I want RVE to become. I want it to be something people can build on with confidence, not a clever demo they have to replace once their product grows. That means listening to the less glamorous problems, making careful changes, and treating real customer projects as the measure of whether the editor works.

The RVE SDK is at the centre of that future. It is delivered through private npm packages that I maintain, so teams can add the editor to their product without taking ownership of the entire codebase. RVE Pro remains there for people who want the full Next.js source, and RVE Cloud Rendering is for the part that should not have to run on a customer's machine. They are different ways of supporting the same goal, not the reason RVE exists.

I do not want to measure the future of RVE by how many features I can add. I would rather make the important parts feel solid, make the path from an idea to a finished video shorter, and give people room to build an experience that belongs to their own product.

There is still plenty I want to improve, and I expect RVE will keep changing as I learn from the people using it. What feels clear now is who I am building for and why the work matters.

I put the first version out to find out whether anyone else shared the problem. I am grateful that people answered. I plan to keep showing up for them.