About

Idler, founded in 2025 and based in San Francisco's Dogpatch neighbourhood, builds reinforcement learning environments that teach AI models to code at expert human levels. Backed by Y Combinator and operating without external venture capital, the company designs training environments based on real-world coding scenarios to prepare AI models for complex production challenges.

The company has generated $440K in revenue and secured a landmark multimillion-dollar contract with a leading foundation lab. Its work sits at the intersection of reinforcement learning, AI model training, and coding environments, serving the AI research, foundation models, and software engineering sectors.

Idler is a small but ambitious team that encodes real human expertise into frontier AI systems, operating on the principle that advanced AI should be trained on rigorous, responsibly produced work grounded in real expertise. The company is rapidly expanding to meet demand that currently outpaces its capacity.

Similar companies

Scale logoSC

Scale

Scale AI is a San Francisco-based data annotation platform that provides high-quality training data and full-stack AI infrastructure to power machine learning models for enterprises, governments, and AI labs worldwide.

1 job
Preference Model logoPM

Preference Model

Preference Model is an AI research company building reinforcement learning environments and data infrastructure to reflect real-world complexity for training machine learning models.

Datacurve logoDA

Datacurve

Datacurve provides high-quality training and evaluation data for frontier AI models, using a gamified bounty platform that engages software engineers on complex coding challenges.

AfterQuery logoAF

AfterQuery

AfterQuery is an applied research lab that creates specialized data solutions, including reasoning traces and agent environments, to support frontier foundation model development.

Weave logoWE

Weave

Weave is an AI-powered engineering analytics platform that helps software teams measure productivity, collaboration, and AI tool impact across their development workflows.

Flexion Robotics logoFR

Flexion Robotics

Flexion Robotics builds the autonomy stack for humanoid robots, creating AI software through simulation and reinforcement learning that enables complex intelligence for simple human tasks.