Merton
An AI project manager for scientists, tracking tasks, decisions, blockers, and artifacts across the messy tools where research actually happens.
Overview
Merton is an AI project manager for scientific teams. It turns the scattered working state of a lab into a living project graph: tasks, strategy decisions, blockers, risks, ownership, and artifact relationships. The system is built for the layer where research teams usually lose context, across Slack, GitHub, Overleaf, runs, uploads, and meeting notes. The goal is not to replace scientists or generate papers. It is to make the research process legible enough that humans and AI collaborators can supervise it together.
Key points
Project Graph for Research
Models strategy decisions, blockers, risks, task ownership, and artifact dependencies as connected state instead of leaving them buried in threads.
Built for Lab Reality
Designed around the tools research teams already use, including Slack, GitHub, Overleaf, experiment runs, uploads, and meeting notes.
Human-Supervised Coordination
Keeps the agent at the coordination layer: surfacing state, flagging drift, and preserving context without mutating code, manuscripts, datasets, or instruments directly.
Source of Bruno
The Bruno paper, accepted at the ICML 2026 AI for Science Workshop, grew out of Merton's core thesis: labs need AI PMs before they need autonomous scientists.
More research
Building AI agents at Sully, which gives health systems more capacity for care: agents for the front desk, the visit, and everything after, connected to the EHR.
Infrastructure that helps the Department of War find and field technology already proven in university research labs.
A constrained, semi-autonomous coordination agent for scientific research teams. Accepted at the ICML 2026 AI for Science Workshop.