Turing

Turing works on both sides of applied AI. For frontier AI labs, it supplies custom and ready-made datasets, reinforcement learning environments, model evaluations and benchmarks. This work covers software engineering, reasoning, enterprise knowledge work and STEM subjects, and draws on a large remote network of engineers and domain experts who rate and correct model outputs.

For large companies, Turing designs and builds agentic AI systems from start to finish. Its published projects include a bilingual Arabic and English AI assistant for a government agency that uses retrieval-augmented generation and natural language queries over databases, enterprise search built on Vespa and Google Vertex AI, and a lending automation platform with a rules engine for a private real estate lender. Other case studies describe coding evaluation suites and reasoning datasets made for model developers.

The company was founded by Jonathan Siddharth and Vijay Krishnan and is based in San Francisco. Companies it shows as clients or partners include Anthropic, Google Gemini, Nvidia and Snowflake. Sectors it names include technology, finance, healthcare and law.