Canvas
Building self-improving agents with continual learning for enterprise AI.
Canvas is an applied-research and product lab focused on building the continual-improvement layer for AI agents in production. It develops post-training, evaluation, infrastructure, and observability tooling that enables enterprise agents to compound improvements over time. The company also produces open-source libraries and research on topics like meta-reward modeling and continual learning.
Canvas provides post-training, evaluation harnesses, infrastructure, and observability tooling that form a flywheel allowing agents to automatically improve from production traces.
Enterprise teams deploying AI agents in production
Background.
- Status
- launched
- Business model
- unknown
- Company
- Canvas
Similar projects.
Editorial take on the space this project sits in — momentum signals, adjacent moves, our call on whether the wedge is real. Get pinged when we publish a new read or when the landscape shifts.
Have a take on this space?
Tell us what you’d build differently, where you think the incumbents miss, or what we’ve gotten wrong about this project. Comments + reactions are coming soon.