Sieve (YC X25) is hiring researchers to build large video datasets for AI labs

Summary

At Sieve, our mission is to push the bounds of how everyone uses video. Video makes up 80% of internet traffic and has become the core component to some of the most consequential industries in the world. Its impact is only accelerating due to AI — think AI-native creative & filmmaking tools, new social & communication mediums, robot systems, open-world video games, and AR/VR glasses. The Video Understanding Problem Generative models power some of the most promising use cases of video: text-to-video generation for creative, talking avatars for communication, world modeling for robot control, interactive video generation for gaming, etc. The core bottleneck to better video models is high quality video data. Video is fundamentally more complex than other modalities and unlike with text, scaling compute to train on the entire internet’s video is intractable due to sheer volume. This means video model quality scales much more effectively with data quality than compute, making data curation and annotation a high priority much earlier on. Solving video understanding removes this bottleneck. Building the most intelligent video AI platform Sieve is a video intelligence company. We build novel video understanding systems with a level of quality and efficiency that empowers developers and enables us to partner with leading video AI model labs to collect, curate, and annotate the highest quality video datasets in the world. Our focus on video means we can invest deeply in video-specific infrastructure and research that improves the iteration speeds of teams working across creative, communication, social, AR/VR, gaming, and robotics by many orders of magnitude — accelerating how these industries transform our lives. Our progress We’ve scaled to millions in recurring revenue as a relatively small team of 11 people, growing 25% month over month. We’re a Series A company and have raised from top-tier firms such as Matrix Partners, Swift Ventures, Y Combinator, and AI Grant. Our cus...