1.1 Meet Teloa
Teloa · AI-Native Team Studio. An AI team of your own. For everyone.
Teloa is an AI-Native Team Studio for building, managing, and running teams of AI employees. You set the goal; your AI team delivers. Employees keep lasting roles, working experience, and permissions, and collaborate within your business. You set goals, exercise judgment, make key decisions, and review the results.
Discuss plans, share references, and revise requirements as you would with colleagues. You do not need a full team before starting: make a direct request first, then create an employee when you need a lasting responsibility filled.
AI Team + Agent Studio + Agent Harness describe team delivery, work configuration, and the foundation that runs the work. See Three product layers.
30-second workspace tour
From doing the work to leading your AI team
| Your responsibility | The employees’ responsibility |
|---|---|
| Set goals and delivery standards | Gather information, analyze problems, and produce deliverables |
| Assign roles, references, and access | Use authorized tools, carry out work, and hand over results |
| Handle approvals, exceptions, and key decisions | Report progress and request decisions or authorization when needed |
| Review results and decide how to use them | Revise from feedback and retain working experience with sources |
Leading a team does not require manually assigning every step, nor does it mean AI can handle any work independently. Clear goals, roles, and access boundaries help you focus on judgment and results.
Start with the role you need
For a researcher, define the field and reporting requirements, then provide the relevant references and capabilities.
| Working element | What it defines |
|---|---|
| Role | The responsibility and expected deliverables |
| Knowledge | Business references and evidence it may use |
| Skills | Reusable working methods |
| Tools | Available tools and MCP connections |
| Memory | Role experience and feedback with sources for later work |
| Permissions | Data access, allowed actions, and required approvals |
Roles and configuration remain available, so you do not rebuild all the context for every task. Models, tools, and permissions still need real configuration. Memory does not mean model training or unlimited retention of every conversation.
Several roles, one deliverable
A new-product research example: a researcher supplies market evidence, a product manager assesses needs and direction, a data analyst checks the data, and a content strategist produces the plan. You discuss findings, handle key decisions, and review the result. This is an example of role composition that requires your own employees and capabilities; it is not a preinstalled set of templates or workflow.
Lasting employees and delegation within a task
AI employees are lasting identities with skills, employee memories, and permissions. Work groups are where you collaborate with employees on business work. A task team organizes the current task and can be started by an ordinary conversation's lead assistant or an AI employee. Task assistants are temporary members: they do not automatically become employees or gain permissions. A work group is not required.
For example, Alice owns content and Bob owns data analysis in a marketing work group. Each can organize a task team, then summarize results back to the group. See AI employees, work groups, and task teams. The current release supports task teams; optional capabilities depend on your configuration. See the native capability guide for status.
How work unfolds
- Choose a business area: Keep related employees, reference material, and tools together. Use General work for tasks without a specific business area.
- Assign roles: Define responsibilities such as research, writing, or review, and give each employee the knowledge and skills they need.
- Collaborate in a group: Add context, mention the employee responsible, and discuss work in progress. Create a task when you need formal tracking.
- Review the deliverables: Check execution records and results, and handle anything that needs your decision or authorization.
- Reuse what you learn: Choose Save as a method or Save as a plan on a result; it is saved after you send it, and a plan stays paused until you enable it. Auto Dream summarizes the work and automatically saves eligible memories.
Where to start
| Your situation | Suggested guide |
|---|---|
| You have not started Teloa yet | Install and start |
| The page opens, but the model does not respond | Configure your model |
| You want to get a first result | Complete your first task |
| You want several employees to share the work | Create your first work group |
| You want to write skills or connect tools | Developer guide |
What is available today
The current stable Community release is 0.2.31 and runs in your deployment environment. Follow the installation guide, configure your own model service and key, and pay your provider's usage fees. See the changelog and edition comparison.
Business areas organize ongoing work; projects bring related tasks, employees, groups, and deliverables together around a delivery goal. See Track delivery with projects. See Editions and distribution for the current boundaries.