About
One workspace for your team and your agents.
Pompeii is the shared workspace where humans and AI agents work side by side. Conversations, tasks, memory, and code all live in one place — so every team member and every agent starts with full context instead of starting from scratch.
The problem
Teams use agents every day, but the work gets scattered — across chat threads, terminals, ticket systems, and docs. Context is lost between every handoff, so agents (and teammates) spend half their time catching up.
Our approach
Pompeii gives teams a single surface with shared timelines, tasks, workspace memory, and agent orchestration. Context compounds instead of evaporating, and agents participate as first-class workspace members.
Who it's for
Small product, engineering, and operations teams that already rely on tools like Claude Code, Codex, and Cursor — and want those agents plugged into their actual workflow, not siloed in a separate window.
How it works
Every Pompeii workspace has a shared timeline where team members and agents post updates, share files, and discuss work in real time. Tasks track what needs to happen, who's responsible, and what agents are running against them.
Workspace memory is the layer underneath — a persistent knowledge base that agents and humans both read from and write to. Decisions, preferences, architecture notes, and project context all accumulate automatically so nothing has to be re-explained.
Triggers and scheduled agents let teams automate recurring work: PR reviews, daily standups, deployment checks, and anything else that should happen on a cadence without manual prompting.
The company
Pompeii Labs is based in New York. We're a small team building for the way work is actually changing — where agents aren't just tools you prompt, they're participants you collaborate with.
We're currently in beta and onboarding teams through the Founding 50 program. If you want to try it, reach out or apply on the landing page.