Skip to main content

How does the Tale platform work?

Tale brings teammates and AI agents into one project workspace. Plan work on task boards, delegate to agents running in sandboxes, and review the results with the context and files that produced them.

How the platform fits together

Plan a project, equip agents for its tasks, and review their results. Knowledge, chat and repeatable workflows support the work your team coordinates.

01 Agents

Agents for your project tasks

Configure project agents with a supported runtime, model and tools. Assign them tasks and follow their progress with your teammates.

Explore Agents

02 Knowledge

Shared sources for the work

Keep manuals, reference files, website content and short facts available for retrieval. Check the cited sources behind an answer.

Explore Knowledge

03 Automations

Delegate multi-step work

Define the steps for recurring work, configure its trigger, and inspect the results of each run. Connector approval policies can pause protected writes.

Explore Automations

04 Governance

Held until someone decides

A protected connector write can wait for approval in an automation run. Review the proposed action, decide, and check its execution outcome.

Explore Governance

05 Arena

Two drafts, one brief

Compare two model responses to the same brief. Read both drafts before choosing which one to use.

Explore Chats

06 Projects

Scope by team

Bring tasks, agents, files and instructions into a project. Choose its audience and explicitly share the chats your teammates should read.

Explore Projects

Platform FAQ

Tale is an open-source project workspace for teams and AI agents. Plan tasks, delegate work, follow progress and review results, with shared context and repeatable workflows.

Chat, Projects, Knowledge, Agents, Automations, and Governance — with Projects scoping work, files, and agents by team.

Yes. Community is MIT-licensed and free to self-host. Enterprise adds support and services on the same codebase.

Start with a project and one task whose result you can judge. Add the relevant files, assign a configured agent, start it, and review what it delivers.

See your team’s next workflow in Tale

Bring a task to a guided demo, or start a workspace and try it with your own agents.