DocETL
A system for agentic LLM-powered data processing and ETL
...The platform allows developers and researchers to construct structured workflows that extract, transform, and organize information from sources such as reports, transcripts, legal documents, and other text-heavy data. Instead of relying on single prompts or ad-hoc scripts, DocETL provides a declarative pipeline framework that breaks complex document analysis tasks into manageable operations that can be optimized and orchestrated automatically. Pipelines are typically defined using a low-code YAML interface, giving users full control over prompts and processing steps while still simplifying workflow creation.