Onehouse, a data lakehouse company, has launched a vector embedding generator as part of its managed ELT cloud service. This new feature aims to automate embedding pipelines for GenAI applications, working with foundation models from OpenAI and Voyage AI.

The vector embedding generator creates pipelines that continuously deliver data from various sources to AI models, which return embeddings to be stored in optimized tables on the user's data lakehouse. It integrates with vector databases for high-scale, low-latency vector serving for real-time use cases while storing all vector embeddings on the data lakehouse.

The vector embeddings generator aims to reduce the time and costs required to build vector embeddings for GenAI applications while providing unique capabilities around update management, late-arriving data, and concurrency control.