Vivek Pandit
Frontier AI Lead - RL Environments · Turing
Building RL environments to improve long range capabilities for frontier AI models
Focus area
RL environments
Systems closest to the work
Evaluations, benchmarks and environments for frontier AI models
Problem being solved
Solving AI agentic experience across various domains
What operating AI in the real world teaches you
Evaluations, taking real feedback from users and building harnesses that adds value orthogonal to what the models provide.
Where AI is creating measurable impact
AI agents for personal use cases seem interesting, falls short on decision making in critical domains
What changes in the next 12–24 months
Infrastructure built for AI agents as the user and consumer