Freesolo
Product Native Models
Y CombinatorSpring 2025ActiveB2BEngineering, Product and DesignArtificial IntelligenceReinforcement LearningB2B
What they do
Problem
Companies need AI models that capture user-trajectory knowledge while delivering better performance than general models at lower latency and cost.
Solution
Freesolo builds specialized product-native models that encode user-trajectory data to outperform state-of-the-art models with reduced latency and cost.
Who it's for
Product and engineering teams at companies building AI-enabled products that want custom, high-performance, low-latency models.
Founders
Eric Mao
Founder
Hello, my name is Eric. I used to study Computer Science and Business at the University of Pennsylvania.
Tom Zheng
Founder
Cofounder @ Freesolo, automating model training.
Company timeline
- Founded2025
- YC batchSpring 2025
- StatusActive
- Team size4 people
- LocationSan Francisco, CA, USA
Similar startups
BentoLabs AI
The monitoring and learning layer for long-running agents
Halluminate
Data and RL environments to automate knowledge work
Coval
Simulation & Evaluation that scales voice and chat AI agents
Refresh
Training gyms for computer use and software engineering work
Exla
An SDK to run transformer models anywhere
Idler
Reinforcement learning environments.