Research

Research and product must be co-designed together with real-world evaluations. Our research culture prioritizes making focused, decisive bets that are driven by real problems sourced from our customers.

Researchers across training, data collection, evaluation, and agent design work together to drive core capabilities in our agentic system.

Prior to Reflection, our team pioneered research in Large Language Models and Reinforcement Learning.

If you're interested in research and product at the frontier of what’s possible in AI, consider joining us.