Docs
Using your workspace

Data

Ask questions of your normalized customer data — every CRM record, call recording, and meeting note — in plain English or SQL. The same data is reachable in the app, over the API, and from an MCP agent.

In the app

Open Datafrom the sidebar. Type a question (“which open deals slipped last quarter?”) and Amdahl writes the SQL, runs it against your unified data, and shows the rows. You can also drop into raw SQL, browse the schema, and drill into pattern clusters — themes Amdahl found across your calls.

What you can reach: a single interactions view that fuses CRM + call + email data, semantic cluster search over recurring themes, and a knowledge-base search across your saved research.

How many open deals do we have by stage?Ask
stageopen deals
Discovery18
Proposal11
Negotiation6

Over the API

The Data Explorer is a read-only REST surface under /api/platform/v1/data. Run SQL directly, or let Amdahl write it from a natural-language question. Authenticate with the X-API-Key header (see Using the API).

bash
# Natural-language question -> Amdahl writes + runs the SQL
curl https://app.amdahl.co/api/platform/v1/data/ask \
  -H "X-API-Key: amdahl_sk_live_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{ "question": "How many open deals do we have by stage?" }'
bash
# Or run SQL yourself against the interactions view
curl https://app.amdahl.co/api/platform/v1/data/query \
  -H "X-API-Key: amdahl_sk_live_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{ "sql": "SELECT interaction_type, COUNT(*) FROM interactions GROUP BY 1" }'

Not sure of the exact route or input shape? Call /api/platform/v1/operations — it returns the catalog of every operation your key can call (each data.* read included) with its JSON schema.

From an MCP agent

Connect Claude, Cursor, or any MCP client (see Using MCP) and the agent gets one data tool with these actions:

  • query — run SQL over the interactions view.
  • explore — read the schema before writing a query.
  • search — semantic search across saved research.
  • cluster_search / cluster_detail — find and drill into recurring themes across your calls.

The agent decides which action to call; you just ask it a question in natural language. Data reads honor the same per-member data-scope rules you set in Settings → Data → Data Access, on every surface — app, API, and MCP alike.