Attio Reports 20x More Agent Tasks

Attio Reports 20x More Agent Tasks
Wednesday, November 18, 2026
  • •Attio says agents now run 20x more customer tasks, one month after its hybrid architecture launch
  • •Ramp’s November spend data names Attio the fastest-growing vendor in the CRM category
  • •More than 35,000 companies use Attio; launch adds SQL, live data connectors, agent scale and public beta Activities

Attio said AI agents now run 20x more tasks for customers, one month after the company launched a hybrid architecture for people and agents. The announcement came as Ramp’s November spend data named Attio the fastest-growing vendor in the CRM category. Examples include Ask Attio creating weekly sales forecast reports and outside agents connected through the Model Context Protocol (a standard for connecting agents to tools) making automatic deal updates.

Revenue teams use agents for account research, lead routing, follow-up, deal updates and forecasting. Attio says agents read context across thousands of records and perform tasks at scale. The company says older CRM platforms were built for people, while its system lets people and agents work with the same live data and rules. CEO and co-founder Nicolas Sharp said agents can query millions of records using SQL while businesses control what agents can access, change and do.

Attio reported a 20x increase in agent tasks since the architecture launch one month earlier. More than 35,000 companies use Attio for go-to-market work, including Granola, Turbopuffer and Modal. Agents already ran in Attio before the launch; the company says they can now handle more complex data tasks, from calculations on a single deal to analysis of an entire pipeline in one query.

The launch added four capabilities: a SQL Engine that supports standard, Postgres-compatible SQL, with queries using the permissions of the sending user or API key; Data Connectors that link BI tools Hex, Looker and Metabase to live Attio data without a warehouse, ETL pipeline or sync delay; Agent Scale engineering for hundreds of millions of records and many agents working at once on the same current data; and Activities, in public beta, which puts growth events, product usage, social messages and other customer signals on the relevant record as structured data.

Subscribe to the AIB Newsletter

The fastest, most reliable AI news, straight to your inbox