HubSpot expanded its Breeze AI platform with two new products: Agent Hub, a central dashboard for deploying and monitoring autonomous agents, and Agent Builder, a low-code interface for creating custom agents connected directly to CRM data. The update was reported by MarTech.org on July 23.
The release shifts HubSpot’s AI offering from prebuilt assistants toward a managed agentic layer inside the CRM, with governance controls built in from the start.
Agent Hub and Agent Builder
Agent Hub gives marketing, sales, and customer support administrators a single workspace to configure, monitor, and manage AI agents. The dashboard consolidates HubSpot’s prebuilt agents for customer support, sales prospecting, and content creation alongside custom agents built by customers. Administrators can review execution logs, track performance across teams, and verify agents operate within company guidelines, according to MarTech.org.
Agent Builder provides the creation side. Teams upload internal documentation, define operating rules, and connect agents directly to CRM records. Agents can use customer records, deal stages, and communication histories to qualify leads, update pipeline data, or create personalized outreach. The low-code interface targets go-to-market teams that want task-specific agents without engineering dependencies.
HubSpot also added operational controls: administrators can set monthly execution limits, monitor credit consumption, and cap individual agent workloads. These cost-control mechanisms address a practical reality for teams scaling agent deployments, where uncapped autonomous execution can generate unexpected compute bills.
Governance as a Feature
The HubSpot product page now positions Agent Hub as the “central home for building and managing AI agents across the platform,” signaling that agent governance is a first-class product surface rather than an afterthought bolted onto existing tools.
This mirrors what enterprise platforms across categories are discovering: autonomous agents need operational oversight from day one. Execution logging, credit caps, and performance monitoring are not premium features. They are table stakes for any organization letting agents act on live customer data.
MarTech Follows Enterprise
HubSpot’s move tracks a pattern visible across enterprise software. Platforms that started with chatbots and copilots are graduating to agent management layers with explicit governance. Salesforce, ServiceNow, and Box have all shipped similar control planes in 2026. The difference is that HubSpot is embedding these capabilities directly into the CRM where customer data lives, removing the integration layer that typically separates agent tooling from the systems of record agents need to access.
For go-to-market teams running agents on live pipeline data, the question is no longer whether agents can do the work. It is whether organizations can see what agents are doing, control how much they do, and shut them down when they go wrong. HubSpot’s Agent Hub is built to answer that question inside the CRM itself.