What changed

LanceDB Inc., a company focused on database technology for AI, announced it has raised $57.5 million in early-stage funding to advance its multimodal data platform. The Series B round was led by Bessemer Venture Partners and Innovation Endeavors, with additional participation from Theory Ventures, Dimension, Swift Ventures, CRV, and Y Combinator. This brings LanceDB’s total funding to $98.5 million.

What this means in practice

LanceDB’s platform aims to simplify the process of training AI models by providing a unified multimodal data lakehouse. This allows researchers, enterprise teams, and AI developers to access, organize, and prepare training data without managing the underlying infrastructure. The company’s technology is built on Lance, an open-source data format that supports structured and unstructured data, including images, video, and sensor data.

LanceDB’s solution is designed to handle the massive scale of multimodal data required for modern AI training. For example, a single training workflow can involve 10 to 20 petabytes of data, with some reaching up to 80 to 100 petabytes.

By keeping data in its original storage location and providing a unified access point, LanceDB reduces the time and complexity of data preparation.

Limitations and availability

The company’s platform connects directly to existing cloud storage, allowing data to remain in place without being copied or moved. This approach minimizes the need for data conversion and network transfer, speeding up the AI development process. LanceDB’s technology is already used by notable customers such as Runway AI, Netflix, and Microsoft. The company plans to use the new funding to improve large-scale data curation, feature engineering, and data loading, while expanding support for emerging AI workloads like video generation and robotics.

Source facts for LanceDB raises $57.5M to expand multimodal data platform for AI training: Bessemer Venture Partners and Innovation Endeavors led the Series B round. AI is making it easier to create and interpret increasingly complex forms of data, but creating it also brings an infrastructure problem. Multimodal AI involves more than combining text, audio, images and video.
Original AINewsTrend explainer based on the cited source. This graphic summarizes source statements; it is not an independent benchmark.

Sources and further reading

  1. siliconangle: LanceDB raises $57.5M to accelerate AI model training with multimodal data platform