Attio ships hybrid agent-native CRM with SQL engine
Attio launched a hybrid architecture for people and agents: SQL engine, live Postgres-compatible data connectors, Agent Scale, and Activities (beta). Queries, dashboards, and agents run against live CRM data with existing permissions.
Attio announced a hybrid CRM architecture built for people and agents in equal measure.
What shipped (beta / plan-gated):
- SQL Engine — agents and workflows can query millions of records and relationships natively inside Attio (Ask Attio on all plans; REST API, MCP, and Workflows SQL on Enterprise).
- Data Connectors — a live, read-only Postgres-compatible endpoint so BI tools (Hex, Looker, Metabase, dbt, and similar) query CRM data without exports.
- Agent Scale — engineered for hundreds of millions of records, with the same permissions and access controls as the rest of Attio.
- Activities (beta) — custom schemas for support chats, product events, and other interactions as first-class CRM data.
Why it matters for GTM engineers: agent stacks need a system of record that speaks SQL and MCP without dumping data out of the CRM. Attio is positioning the CRM itself as queryable infrastructure, not just a UI over Salesforce-style objects.
Sources: Attio blog, founder announcement.