AI Scrum Master · Private beta

Run the standup. Leave with accountable action.

huzlr joins eligible Google Meet standups as an identified AI participant, facilitates with your team's instructions, and turns the conversation into a reviewable delivery record.

Standup output

Tuesday delivery sync

Completed

Summary

The preview environment is ready. Release validation moves to today.

Blocker

Production API credentials are still awaiting approval.

Action

Maya will confirm credentials before the release review.

Illustrative output. huzlr preserves unknown owners or dates instead of filling gaps with guesses.

How it works today

A controlled path from calendar event to delivery record.

The workflow is designed around explicit workspace policies, visible participation, and outputs your team can inspect.

01

Connect the meeting calendar

Use a dedicated Google account and sync the workspace calendar so huzlr can identify eligible Google Meet events.

02

Set an explicit join policy

Choose manual or automatic joining, limit which meetings qualify, and set a maximum meeting duration.

03

Run the standup

The named AI participant joins the call, introduces itself, listens, and facilitates using your workspace instructions.

04

Review the output

Keep the transcript when enabled, then review the summary, decisions, blockers, and proposed action items.

Controls before automation.

A meeting agent participates in a sensitive team ritual. huzlr makes the join policy, identity, and retained outputs visible workspace choices.

  • Auto-join is opt-in and can be restricted to explicitly tagged meetings.
  • The agent joins under a visible bot identity and announces that it is an AI meeting assistant.
  • Transcript, summary, and action-item outputs can be enabled by workspace.
  • Meeting duration and join timing are bounded by workspace settings.

Clear limits build trust

huzlr is designed to state uncertainty instead of making the meeting record look more complete than the conversation was.

  • Participant names are not inferred when the mixed meeting audio cannot support reliable diarization.
  • Owners, due dates, decisions, and blockers should not be invented when the meeting did not establish them.
  • Outputs are proposed coordination artifacts for human review—not automatic proof that work is complete.

Questions before joining

What teams ask first.

Does huzlr replace the human Scrum Master?

No. It handles repeatable meeting coordination and drafts the record of the conversation. Teams still own judgment, facilitation choices, and delivery decisions.

Can the AI Scrum Master join every Google Meet automatically?

Only if a workspace administrator enables auto-join. Teams can require an explicit Scrum tag, limit eligible organizers, and configure timing and duration controls.

What does huzlr create after a standup?

Depending on workspace settings, huzlr can retain the transcript and generate a concise summary with decisions, blockers, and proposed action items.

Is the AI Scrum Master generally available?

It is currently offered through the huzlr private beta. Availability and onboarding depend on the meeting environment and workspace setup.

Private beta

Make the standup produce movement, not another round of status chasing.

Create a workspace to request access, or read the huzlr engineering blog for more on meeting-to-action systems.