#synthetic-data

Kiln is a free tool for building production-ready AI systems. It supports RAG pipelines, evaluations, agents, MCP tool-calling, synthetic data generation, and fine-tuning. - Seamlessly connect agents and tools through MCP servers/interfaces - Build, fine-tune, and run models locally or remotely - Generate synthetic data, orchestrate evaluations, and manage datasets - Collaborate across technical and non-technical stakeholders via Git-based workflows - Use a free, open source Python library to embed Kiln into your own systems - Maintain data privacy: Kiln runs locally; external APIs are called only via your own keys **Website / Docs:** https://kiln.tech

Generate realistic relational test data in one API call. 156 field types, 22 locales, foreign key references across tables, JSON/CSV/SQL output. Seed your database with production-quality fake data.

Simulate and evaluate voice & text agents from your editor. Generate synthetic-user "Drivers" from your codebase, run multi-turn simulations against your agents, and pull back transcripts, traces, and scores without leaving your AI assistant.