New architecture enables AI agents to analyze millions of records, automate sales workflows and work alongside revenue teams using live CRM data
Attio, a platform for agentic revenue operations, announced that AI agents are now performing 20 times more tasks on its platform, just one month after the company introduced its new hybrid architecture.
The tasks range from generating weekly sales forecast reports through Ask Attio to automatically updating deals using external agents connected through the Model Context Protocol (MCP). The increase reflects growing adoption of AI agents for core go-to-market activities, including account research, lead routing, follow-ups, deal management and forecasting.
The announcement comes as Ramp’s November spend data identifies Attio as the fastest-growing vendor in the customer relationship management (CRM) category.
AI Agents Take on More Revenue Operations
Attio said its updated architecture allows AI agents to work with increasingly complex datasets and execute tasks at scale alongside human users.
While agents previously operated within the platform, the new architecture enables them to handle more demanding workflows, from performing calculations on individual deals to analyzing entire sales pipelines through a single query.
More than 35,000 companies use Attio for their go-to-market operations, including Granola, Turbopuffer and Modal.
Nicolas Sharp, CEO and co-founder of Attio, said traditional CRM platforms were designed primarily for human users, whereas modern revenue teams increasingly need systems that support both people and AI agents.
He explained that agents can query millions of records using SQL while businesses retain control over data access, permitted changes and the actions agents can perform.
Four Platform Enhancements Support Agentic Workflows
Attio’s new architecture introduces four capabilities designed to improve data access, scalability and automation.
SQL Engine: The platform’s system of record can now be queried using standard, PostgreSQL-compatible SQL. This allows AI models to interact with CRM data through a widely recognized query language. Queries operate under the permissions of the user or API key initiating them.
Data Connectors: Business intelligence tools, including Hex, Looker and Metabase, can connect directly to live Attio data. The integrations eliminate the need for a separate data warehouse, ETL pipeline or delayed synchronization process.
Agent Scale: The updated infrastructure supports hundreds of millions of records and multiple agents working simultaneously, with all agents accessing the same current data.
Activities (Public Beta): Customer signals such as growth events, product usage and social messages can be added to relevant CRM records as structured data. Both employees and AI agents can use these signals to inform subsequent actions.
External AI Agents Connect Through MCP
The platform also supports external agents through MCP, allowing connected AI tools to perform tasks such as automatic deal updates using Attio data.
By providing agents with access to live customer relationship context and existing business rules, the architecture is designed to support more coordinated revenue workflows while maintaining organizational controls.
Attio said the combination of structured data, SQL access and permission-based operations enables agents to take on more complex revenue activities than were previously practical.
Independent Spend Data Signals CRM Growth
Market data also points to increasing momentum for Attio. Ramp, which tracks vendor spending across more than 85,000 U.S. businesses, identified the company as the fastest-growing vendor in the CRM category in November 2026.
Attio was also the only CRM provider included in Brex’s Fall 2026 list of the 25 fastest-growing software vendors, ranked by growth in customer spending.
The latest developments reflect a broader shift toward agentic revenue platforms, where AI systems work directly with live business data to support sales teams, automate operational tasks and improve visibility across customer relationships.
