Eventual

Daft is an open-source data engine designed to power LLM-driven data pipelines by transforming raw multimodal data into vectors, labels, and structured outputs. It provides a unified framework that combines ingestion, chunking, embeddings, and multimodal transforms, offering built-in scaling and orchestration to eliminate the need for infrastructure management. By enabling reliable model-on-data pipelines across millions of rows, it serves AI engineers and data teams who need consistent behavior from local development to production. The tool is accessible as an open-source project on GitHub and offers a serverless platform for those requiring zero infrastructure headaches.