Role-based access
Tenant roles and per-project permissions apply to people and AI roles alike. The web app, the API and the AI tooling all consult the same checks.
AI Project Operations
EtherPM brings people, specialist AI roles, decisions and delivery into one operational workspace — so every project has direction, ownership and momentum.
Projects, tasks, discussions and documents sit in the same place, so context never has to be rebuilt.
Specialist AI roles work alongside people, under the same roles and the same permissions.
The discussion, the decision and the work it changes stay attached to the project that owns them.
Sprints, roadmap and Kanban show what is moving, what is waiting and what is at risk.
Platform
EtherPM is one operational workspace: the plan, the people, the AI roles, the conversation and the delivery signals all read from the same project.
Epics, tasks, sprints, roadmap and Kanban stay in step with the work itself. A project keeps a current state instead of a document that ages quietly.
AI Architect, AI Developer, AI Tester and AI Project Manager join a project the way colleagues do — with a defined remit and work that lands in the project.
Project discussions, comments, shared workspaces and invitations keep the conversation next to the work it concerns, for everyone with access to it.
A knowledge base and project AI memory give people and AI roles the same history to work from, so a question answered once stays answered.
Overdue work, blocked tasks and open issues surface as project signals rather than as a status report somebody has to assemble by hand.
How it works
Four steps, and a person decides at each of them what the project actually takes on.
Describe the outcome you want. AI Architect proposes a structure — epics, tasks and a first plan — which you review before anything is created.
Invite people, bring in the AI roles the project needs, and set what each of them may see and change.
Work moves through sprints, boards and discussions. Ownership and progress stay visible without anyone chasing an update.
Deadlines, blockers and open issues are raised early, with the context needed to decide and move on.
Control and governance
Every AI role works inside the same permission model as the people it works with, and its output arrives as project content a person can read, change or discard.
Tenant roles and per-project permissions apply to people and AI roles alike. The web app, the API and the AI tooling all consult the same checks.
A generated plan is previewed and confirmed by a person before it becomes epics and tasks in a project. Nothing is created behind your back.
Activity feeds and audit logs record what changed, who changed it and when — for human and AI actions in the same record.
AI work lands as tasks, comments and documents you can open and judge. There is no silent change to review after the fact.
EtherPM can talk to an Ollama server on your own network, so project context can be worked on with a model you host yourself.
Workspaces, project sharing and invitations decide who joins a project, at what level, and for how long.
EtherPM is in early access. Tell us how your projects run today and we will show you how the workspace fits around them.