# Connect imported conversations to the right People.

v0.133 · Released 2026-08-24

Conversation imports can add known external attendees to People and connect their conversations, Studies, votes, and portal activity. Uncertain matches remain unassigned.

## What changed

- Add known external attendees to People during conversation imports
- Leave uncertain Account matches unassigned
- Review conversations, Studies, and votes around the same Person
- Choose People and Account behavior in the Signals API and MCP

## Keep customer context with each conversation

Imports from Granola, Fireflies, Fathom, Gong, and Microsoft Teams can add known external attendees to People. The import result shows which People and Accounts were created or connected before the team continues.

- [Read the People and Accounts guide](/docs/product/people-and-accounts)

## Use the same Person across research and feedback

People is the workspace directory for customers and research participants. A Person can belong to an Account or remain unassigned. Portal activity connects to an existing Person only through an exact email match.

- [Import customer evidence](/docs/product/evidence-to-insights)
- [Use the Signals API](/docs/developers/signals-api-quickstart#organize-people)

## More in this release

### People and Accounts

Customer identity stays useful without turning every attendee into CRM data.

- Added a workspace People directory backed by existing Contacts, with optional Account assignment
- Connected People to conversation, Study, and vote activity while preserving historical participation
- Kept internal, unknown, ambiguous, and email-less attendees from creating noisy People records

### Imports and API

Every supported intake path uses the same deliberate identity rules.

- Added separate controls for saving customer attendees and creating missing Accounts
- Added People and Account outcomes to conversation import receipts
- Added participant policies, stable source identity, and explicit Account creation to the Signals API and MCP

Canonical page: https://zentrik.ai/releases/v0-133