Abhinav Soni — Co-Founder

Building Bento Labs (YC P26). The infrastructure to improve and monitor your AI agents. We are building the monitoring that understands what to catch, and does not wait for you to add a regex pattern to wait for. We are building the evals that don't go stale, because no one maintains them. We are building the learning layer that actually completes the loop of agent improvements, that makes run 101 significantly better than run 1. Previously led AI agents at Emergent through 0 → $100M ARR. Spent two years reading hundreds of production trajectories a day. Improved the harness, every single day. What I learned: the model is not the bottleneck. The harness that encompasses it is. PS: some people called me "the agent whisperer." :p Would love to talk about: anything around AI agents. I write about agents, prompts, evals, and what two years of reading trajectories taught me. Follow if any of that is useful.

Stackforce AI infers this person is a SaaS and AI specialist with a focus on product management and autonomous systems.

Location: San Francisco, California, United States

Experience: 2 yrs 11 mos