YC-Backed Dumme Raises $3.4M for Its AI Video Editor That Turns Long-Form YouTube Videos into Short

Dumme, a startup putting artificial intelligence into video editing, is already generating demand before opening to the public. The Y Combinator-backed company has hundreds of video creators testing its product, which leverages AI to create short-form videos from YouTube content, and a pre-launch waitlist of more than 20,000, it says. Using a combination of both proprietary and existing AI models, Dumme’s promise is that not only can it save editing time, but also and here’s its big claim to do a better job than the contracted (human) workforce that she is often in charge of menial video editing. jobs, like trimming long-form content for posting on short-form platforms like YouTube Shorts, TikTok, or Instagram Reels.

Founded in January 2022 and a participant in startup accelerator Y Combinators’ Winter 2022 program, Dumme co-founder and CEO Merwane Drai said he was initially focused on building a search engine for video. But about six months ago, the team realized that a better product might be to reuse the same AI models they were developing to edit video clips instead.

Along with co-founders Will Dahlstrom (CPO) and Jordan Brannan (CTO), all with backgrounds in AI, Drai realized Dumme may have landed on the right product for the market after their app went viral. blocking their servers.

We didn’t really expect it to have a lot of traction or anything, so we just put something out there, Drai explains. So what happened was that overnight we woke up to overloaded servers like nothing was really working. So we took everything apart and put together a sort of waiting list, he continues. The next morning, we woke up to probably 5,000 people in there, which was interesting.

The team later discovered that a TikTok creator had posted a short video about the product, which sent a surge of traffic to their site.

He never really calmed down, notes Drai.

The product, pronounced fictitious, liked the creators because it aimed to simplify and speed up the work related to video editing.

Using Dumme is as simple as the name suggests. To get started, the user pastes a YouTube video link, then clicks Generate, and the AI ​​will produce a series of short videos showcasing the highlights of that imported content. The company says it is using YouTube as a source, instead of supporting raw video footage, in order to outsource content moderation which, while allowed on YouTube, is allowed in Dumme.

Processing time and the number of resulting clips will depend on the length of the original video.

But for example, an hour-long video podcast might take about 20 minutes to process, and you’ll start getting clips after about five minutes, says Drai. Once done, creators can download the video clips, which by default are under 60 seconds, and upload them to any platform that supports short-form content, like YouTube Shorts, but also other platforms, like Reels or TikTok.

How all of this works on the backend is, of course, much more complex. The company says that, initially, Dumme will learn as much as possible from the source video via metadata. Then he transcribes the video and tries to understand the semantics of what is being said by also looking at the frames to try to decode the emotions of the person speaking. These results are correlated and passed to a language model which tries to determine which parts of the video are worth trimming. This is then passed to another model which tracks active speakers and handles clipping.

Dumme says he’s working with existing AI models like GPT-4, a refined version of Whisper, and others he’s built in-house like the model that tracks active speakers in a video frame. One of her models is even trained on a series of YouTube shorts to learn what makes a good opening hook to attract viewers. And, while it’s not live yet, the team is also experimenting with an open source model, LaViLa from Facebook Research, to better understand the context of the video.

The AI ​​work is done on the Cloud GPU provider CoreWeave, not AWS, as it is more cost effective, the company tells us.

Because Dumme relies on AI processing spoken words, the technology isn’t appropriate for things like long gameplay videos or others where people aren’t talking. Drai says the startup is initially targeting YouTube creators, podcasters and agencies, which they believe would be the best fit to monetize the product.

Agencies, Drai explains, today often outsource this type of work with spot-on results.

They only pay contractors in cheap jurisdictions to edit their content. And the problem is, it’s actually still quite expensive and it takes a long time, it takes weeks, not minutes, he says.

Asked how he feels about creating a technology that would actually put people out of work, Drai wasn’t concerned.

The way I think about it is, in the end, I think it’s like telling me that math teachers will do it [be put] out of work because there’s something called a calculator, she explains. People will adapt. And then there will be someone who will teach you how to use the calculator, right? So I think it’s just a matter of adjusting to that, Drai says.

Currently, the pricing under consideration has tiers where a business would pay $0.40 per minute of processed video, while smaller creators can instead opt for a monthly subscription limited to 10 hours of content per month. (These numbers may change.) During testing, the product was used free of charge.

Early adopters have used Dumme for a variety of edits, including generating clips from their video podcasts to post on Shorts, as well as trimming other new videos and examining their previous catalogs.

The product appears to be competitive with other AI technologies on the market, including that of creator company Jellysmack, which has leveraged AI to turn longer YouTube videos into shorter videos by trimming, resizing and optimizing them for specific platforms . of its acquisition of Kamua in 2021. Other tools that do similar work include things like Opus Clip, Vidyo.ai, Detail, TubeBuddy, Wisecut, and others. How far Dumme will succeed or fail will be in outperforming competitors on work quality and cost metrics that have yet to be determined.

But some investors are betting on Dumme. Ahead of launch, the startup raised a $3.4 million seed round from Y Combinator, Caffeinated Capital, Max and Nellie Levchin (through SciFi VC), Suhail Doshi, Nico Chinot, Protocol Labs, Chris Puscasiu, and other angels.

Given the interest and sizable waiting list, Dumme says he aims to get around 500 people on board each week. TechCrunch readers can skip the line by using the invite code TECHCRUNCH until slots run out.