Meet trAI, your in-meeting AI assistant

trAI is the TeamRetro AI assistant that safely and privately analyzes and summarizes your meeting, suggests actions and acts as a co‑facilitator for your session.

Try TeamRetro free
The trAI side panel opening in a retrospective. It describes itself as a private assistant that only ever sees what you can see, and offers three starting prompts: 'Summarize the key themes so far', 'Draft actions from top-voted ideas' and 'What's still open from last time?'

The team’s intelligence, in the room with you

trAI sits in the side panel while you facilitate, ready when you need more than what’s on the board. Ask it to make sense of where the discussion has landed, to sharpen the action the team just agreed on, or to surface what they said about this three retros ago. It’s recall no one in the room has.

It works in retrospectives, health checks and estimation sessions, and acts only when you ask it to.

What can your AI meeting assistant do in TeamRetro?

Generates new formats

Get fresh new ideas for retro formats, health checks and icebreakers.

Summarizes as you go

Surface key themes during or at the end of meetings for you and the team.

Gives clever suggestions

trAI proposes unbiased actions, suggests next steps, and flags what you’ve missed.

Acts on your instruction

Use natural language instructions for trAI to act on your behalf, saving you time.

Recalls past outcomes

trAI helps your team with follow through without having to dive into the archives.

Spots the patterns

Get more insights beyond the meeting to track hot topics, themes and team sentiment over time.

trAI and the MCP server

Two doors, and the difference is where you’re standing when you ask.

trAI inside the meeting

For the person facilitating right now. It lives in the side panel of a retro, standup, health check or estimation session, and works on the meeting in front of you.

MCP outside the meeting

For the rest of your week. The TeamRetro MCP server connects Claude, ChatGPT, Copilot or Gemini, so you can ask from the assistant you already work in.

Most teams end up using both: trAI while the meeting is running, MCP for everything between meetings.

Three things to try this week

In a retrospective

  • What are the themes so far?

Reads the board as it stands and hands back the shape of the discussion, so you can check the team agrees before moving on.

In a health check

  • Which dimension needs focus?

Puts this session next to the previous one and points at what moved, while the radar is still on screen.

In an estimation session

  • How have we sized similar work?

Looks up comparable past sessions, so a contested estimate has some history behind it rather than just opinions.

Enterprise-grade security, compliance, and control

TeamRetro is SOC 2 Type 2 accredited by an independent auditor who has evaluated our policies, product, platform, and infrastructure in accordance with the Standard on Assurance Engagements (ASAE 3150).

REQUEST SOC 2 VIA TRUST CENTERDOWNLOAD SOC 3
SOC 2 Type 2SOC 3

  • GDPR compliance and regional hosting (US & EU)

  • Single Sign-On (SSO) via SAML

  • SCIM for scalable user provisioning

  • Audit logs available on request

We maintain 99.99% uptime, with fast-response SLAs to ensure your teams stay productive and supported.Status dashboard

Frequently asked questions

What is trAI?
trAI is the TeamRetro AI assistant. It sits in the side panel of a retrospective, health check or estimation session, so you can ask for a summary, capture an action or look something up without leaving the meeting you're running. trAI acts as a co-pilot for your Scrum Master to help them and the team focus on the meeting topic itself, rather than the admin. It's the in-meeting half of TeamRetro's AI: the MCP server is the same data from the assistant you already work in, outside the meeting. For how it works step by step, see trAI in the help centre.
Is trAI included in our plan?
Yes. trAI is in beta and included in all plans, and it can be turned on or off on any of them. Rate limits apply. On Enterprise plans, admins can manage AI settings for their account.
Can trAI be wrong?
As with any language model, AI can make mistakes. While it can read the text, it cannot read the room. Judgment and checking outputs are still required.
Is our data used to train AI models?
No. Your meeting content is never used to train a model — not by us, and not by our providers. TeamRetro doesn't build or train models at all. When you ask trAI something, the relevant content is sent to AWS Bedrock to answer that question, and Bedrock doesn't log it, store it, or use it for anything else. The same holds for every AI feature in TeamRetro, not just trAI. The full picture is on responsible AI.
Which AI models do you use, and who provides them?
Claude models run the in-meeting work — trAI, summaries, grouping, action suggestions and insights — through AWS Bedrock. OpenAI's ChatGPT generates templates, health models and icebreaker questions. Whichever feature you're using, the output is a suggestion: a person reviews it and decides what to keep. The full table naming the provider behind every AI capability, and the model wherever we run a named one, is on responsible AI.
Where is our AI data processed?
TeamRetro runs two fully isolated production environments, one in the US and one in the EU, and your data stays within whichever region you choose. The US environment is hosted in Northern Virginia, and the EU environment in Frankfurt, Germany. You select your hosting region when you create your account, and trAI processes data within that same environment, so nothing crosses regions. If you need to switch regions later, reach out to info@teamretro.com.
How long is our data kept?
The model keeps nothing — AWS Bedrock doesn't log or store what trAI sends it, so no prompt history builds up on the provider's side. What trAI creates in the meeting, such as an action, an idea or a comment, is ordinary meeting content: it lives in your account like anything else the team wrote, and follows the same retention. It stays while your account is active, is deleted 365 days after a trial or subscription ends, and is removed within 60 days of a deletion request. The full policy can be found in security.
Is trAI covered by your SOC 2 report?
TeamRetro is SOC 2 Type 2 accredited for security, availability, confidentiality and privacy, covering our policies, product, platform and infrastructure — trAI runs inside that platform rather than as a service bolted on beside it. For the current report and exactly what the latest audit period covers, request it through our Trust Center. The AI-specific controls, including the AWS Bedrock guardrails we run over inputs and outputs, are described on responsible AI.