What each connection unlocks
Each layer builds on the one before it. Some connections need administrator approval in your organization. We are here to help make that happen.
The core context stack
Your AI platform
Sources: Claude, ChatGPT, or any other client through the MCP connector. This is where your sellers talk to Tempo. The MCP connector exposes purpose-built tools: get deal state, search cited observations, list meetings, review pending CRM updates, run a Skill. Ask Claude to prep you for your 2pm, and the answer is grounded in your actual deal.Your CRM
Sources: Salesforce, HubSpot, or a CSV import. Your CRM tells Tempo what your deals are. Tempo ties every call, email, and conversation to the deal it belongs to, proposes CRM updates with citations back to the source, and keeps a real-time picture of each deal’s state.How Tempo writes to your CRM is set per field. Manual generates no proposals, Suggest holds a proposal for review before anything is written, and Automatic applies the update without review. See Fields and CRM objects for the full rules, including what applies without review.
Call transcripts
Sources: Granola, Gong, Fathom, and Google Meet transcripts through Google Workspace. Transcripts hold the exact words your customer used and the commitments your champion made. Tempo pulls them in, extracts observations, and links them to the right deal.Email and calendar
Sources: Gmail and Google Calendar. Email carries the parts of a deal that never make it onto a call: procurement threads, security reviews, pricing clarifications. Calendar tells Tempo who you are meeting and when.Conversations and docs
Sources: Slack and Notion. Slack channels linked to a deal bring in the day-to-day conversation around it. Any member can connect Slack reader access to link private channels and group DMs to deals they own. Notion pages linked to a deal bring in deal plans, account notes, and handoff docs.Warehouse
Sources: BigQuery. If your team already replicates call and CRM data into BigQuery, Tempo can read it from there with a read-only connection.How Tempo treats your data
- Read-first by default. Tempo reads from your systems to build context. The writes it makes back to your CRM follow the update behavior set on each field.
- Your data is yours. We do not train foundation models on it.
- Scoped access. Each integration page lists the exact scopes Tempo requests and what it does with them.
- Tempo does not send email. Tempo reads mail to build context. It has no path that sends mail on your behalf.
Ready to connect
Context sources
Salesforce
Sync opportunities, accounts, and contacts from Salesforce.
HubSpot
Sync deals, companies, and contacts from HubSpot.
CSV import
Upload your opportunities directly if you cannot connect a CRM yet.
Google Workspace
Connect Gmail, Calendar, and Meet transcripts.
Granola
Bring in meeting notes and transcripts from Granola.
Gong
Pull call recordings, transcripts, and Spotlight AI content from Gong.
Fathom
Pull meeting recordings, transcripts, and AI summaries from Fathom.
Slack
Capture conversation context from channels tied to your deals.
Notion
Sync page content linked to your opportunities.
BigQuery
Read call and CRM data replicated into your warehouse.
AI platforms
Claude
Talk to Tempo inside Claude for deal-grounded answers.
ChatGPT
Use ChatGPT as your interface to Tempo.
MCP connector
Connect Tempo to any AI platform over Model Context Protocol.

