MarkosWebWeb intelligence for the agentic economy

Polymath

ActiveAI-native

Simulation environments to train & evaluate long-horizon AI agents

polymathlabs.aiAI readiness reportLinkedIn

What it does

Polymath is a data lab focused on improving long-horizon agent capabilities and alignment

Next to it in Agent infrastructure · Simulation environments for agent training

  • Ooak DataY Combinator S26

    We turn company data into training data

  • VerisPlug and Play PnP 2026

    Veris is a sandbox platform that lets enterprises train and validate autonomous agents in realistic, high-fidelity simulations before deployment.

  • Ressl AIY Combinator W26

    Train, eval and build autonomous agents

  • Arga LabsY Combinator X26

    Real-world sandboxes to test and train AI agents

  • hiloopY Combinator S26

    Infrastructure for recursive self-improvement

  • MangoDeskY Combinator S25

    Long-horizon RL environments

  • AviroY Combinator X25

    Environments for Long Horizon Tool Use

  • RefreshY Combinator X25

    Training gyms for computer use and software engineering work