Sieve

Sieve builds data infrastructure for AI labs, sourcing, filtering, indexing, annotating and delivering video, audio, image and interaction data at scale. The company describes itself as the only AI research lab focused exclusively on video data. Its video infrastructure operates at exabyte scale, and it has indexed billions of videos, images, audio clips and interaction traces using purpose-built detectors and embeddings, adding dense labels, pairings, temporal alignment and human QA.

A second pipeline captures multimodal data across real-world, digital and simulated environments and scores it for semantics, rights, artifacts and task quality. The technical work spans video understanding, video indexing and embeddings, computer vision detectors, data annotation and labeling, temporal alignment, semantic scoring and data quality, and AI training data curation. The company partners with top AI labs and reports significant revenue; it is at Series A stage.

Sieve is based in San Francisco and works in person at its SF headquarters. The team numbers 15 people, with engineers holding direct ownership over projects end to end. Stated expectations include writing clean, maintainable code, moving quickly without creating brittle systems, strong communication skills and a bias to action, alongside a deep interest in video, media technologies and frontier AI applications.