Show HN: LocalFlow – AI data analysis where the LLM never sees your data

Category: library

Tags: privacy, data-analysis, ai-framework

Score: 5.8/10 (Innovation: 6, Technical: 6, Documentation: 6, Utility: 5)

LocalFlow is a metadata-first AI framework that enables AI-driven data analysis without exposing raw data to the LLM, using schema and statistics for code generation that runs locally. It offers a novel approach to privacy-sensitive data analysis with deterministic, repeatable results, but is still early-stage with limited adoption.

Target audience: backend devs

Repository: https://github.com/localflow-ai/localflow-core · TypeScript · NOASSERTION · 5 stars

View on Hacker News