Activeloop
Activeloop builds Deep Lake, an open-source, Postgres-compatible database for AI that indexes, searches and streams large multimodal datasets to GPU training pipelines.
Activeloop, founded in 2018 and a Y Combinator 2018 batch company, builds Deep Lake, a Postgres-compatible database for AI. Deep Lake indexes, searches, organises and reasons over large multimodal datasets - documents, images, video, LiDAR and telemetry - using natural language and LLM-powered workflows, and streams the data to GPU training pipelines. The software is open source, hosted at github.com/activeloopai/deeplake, and has attracted 9k GitHub stars and 5m downloads, alongside a Slack community of more than 3,200 members.
The company's technical work spans AI data infrastructure, multimodal databases, retrieval augmented generation, natural language querying, and data streaming for computer vision, LiDAR and time-series data. Deployments run in industries including MedTech and radiology, biotech, legal, and research and academia; reported results include an 18% improvement in retrieval accuracy for RAG at Flagship Pioneering and an 18.5% productivity gain for lawyers using Hercules.ai at Ropers Majeski. Deep Lake is in production with organisations including Bayer Radiology, Yale, Mila Quebec and Oxford, and the company works with Intel through the Intel Disruptor Program and with AWS via the AWS Partner Network and AWS Marketplace.
Activeloop is headquartered across hubs in California and Yerevan, with global enterprise deployments, and prefers onsite collaboration in both locations. It describes a fast-moving culture built on shipping, ownership and solving hard technical problems, with structured interviews using objective criteria and merit-based hiring as an Equal Opportunity Employer. The company publishes salary ranges on every role, reviews compensation annually, and offers meaningful equity, medical, vision, dental and life/AD&D insurance, and a 401(k) plan, together with mentorship and learning support.
The company has raised just shy of $20m in total, including an $11m Series A, from investors among them Streamlined Ventures, Y Combinator, Samsung Next, Alumni Ventures and Dispersion Capital.
