Halluminate
Data and RL environments to automate knowledge work
What they do
Computer- and browser-use AI agents are inaccurate, slow, and costly because organizations lack high-quality evaluation datasets and realistic sandbox environments for training and testing.
Halluminate provides proprietary datasets with expert annotations plus realistic simulated software environments to benchmark, evaluate, and train computer-use AI agents with reinforcement learning.
Foundation model labs and enterprises building and improving AI agents for knowledge work.
Founders
Jerry Wu is co-founder and CEO of Halluminate. Previously, he led product and research at Capital One Labs, launching one of the first AI agents in financial services and co-authoring three patents. Jerry studied Computer Science and Economics at Cornell, where he researched model quantization methods and served as VP of the Cornell Consulting Group. He was also class speaker at Acton-Boxborough High School.
2x early startup data/sw eng Building a platform for environments, evals, and benchmarks to train and test computer agents AMA about LLM/agent benchmarking and evals! https://halluminate.ai/
Open roles 1
Company timeline
- Founded2024
- YC batchSummer 2025
- StatusActive
- Team size5 people
- LocationSan Francisco, CA, USA
Similar startups
