HubSpot's Fall Spotlight 2026 updates make one thing clear: Breeze Assistant is becoming much more than a place to ask questions.
HubSpot is bringing Breeze deeper into the work teams already do across the CRM. It can maintain context across related conversations, help build agents and workflows, work with CRM data, create reusable outputs and increasingly take action instead of simply returning an answer.
That shift is what I'm paying attention to.
The interesting question isn't whether HubSpot added more AI features. It's whether Breeze can help a team get to an outcome faster without creating another tool people have to remember to use, another place to copy information or another AI conversation where you have to explain your business from scratch.
Projects give teams dedicated workspaces for ongoing AI work, bringing together related conversations, shared instructions, reference materials and HubSpot context so Breeze doesn't have to start from scratch every time.
Scheduled Prompts let teams set Breeze Assistant prompts to run automatically on a recurring schedule. Instead of remembering to ask Breeze for the same analysis or update every week, teams can make recurring AI work part of the process.
Agent Builder lets users describe an agent's job in plain language. Breeze Assistant can propose a configuration including goals, tools and knowledge sources, then open it in Agent Builder for testing and refinement. Agent Builder is in public beta.
Workflow building brings Breeze into Agent Hub's workflow builder, where it can help configure triggers and actions and generate code for custom-code actions.
Artifacts turn Breeze output into usable content that can continue to be opened, shared and refined rather than disappearing inside a chat.
Voice Mode (Public Beta) brings Breeze into a more conversational mobile experience, where users can ask about CRM data, update records, create follow-up work and complete other actions using natural language.
Individually, those are useful features. Together, they point toward something bigger.
Sam Marks came out of HubSpot Partner Day with three words for our team: Context. Context. Context.
After the Fall Spotlight keynote, I think that's one of the most important ideas for HubSpot customers to understand.
AI can generate an answer without knowing much about your company. But if you want AI to participate meaningfully in your business processes, it needs to understand more: your customers, terminology, processes, CRM data, trusted knowledge, desired outcome and the actions it should be allowed to take.
HubSpot's current context tools include Projects and knowledge vaults, which can make business knowledge and CRM context available to projects and agents. Knowledge vaults are currently in Beta.
AI doesn't make clean data, connected systems and good processes less important. It makes them more important.
We'll be taking a deeper look at HubSpot's context layer as part of our ongoing Fall Spotlight 2026 analysis.
How often do you start a new AI conversation by explaining the same company, audience, process, source material and goal again?
Breeze Projects give teams dedicated workspaces for related AI work. A Project can carry shared instructions and reference materials across conversations, and teams can also bring in HubSpot context such as CRM records and knowledge vaults. Depending on the configuration, Projects can use tools that allow Breeze to do more than generate an answer, including working with connected information and taking permitted actions in the CRM.
Think about a quarterly pipeline review, campaign planning, customer account reviews or a recurring sales process. Instead of rebuilding the brief every time, the Project can carry more of that context forward.
That starts to make Projects more interesting than simply organizing AI conversations. The goal is to create a working environment where Breeze has more of the context, information and tools it needs to help with an ongoing business process.
But there's an important flip side. If the instructions are unclear, the files are outdated or the underlying CRM data is unreliable, AI can carry that context forward too. Persistent context is only as useful as the context you're persisting.
Agent Builder is one of the Fall Spotlight capabilities I'm most interested in from a technical perspective. Lowering the barrier to building agents is exciting, but it doesn't eliminate the hard part: deciding what the agent should actually be responsible for.
Making agents easier to build makes those questions more important, not less important.
That's also where our work around AI agents and assistants fits. Building an agent is one part of the job. Defining the right responsibility, context, systems, actions and human oversight around it is what turns it into a business solution.
HubSpot's Agent Hub workflow documentation shows Breeze Assistant helping users add workflow triggers and actions and generate code for custom-code workflow actions.
AI isn't replacing the need to understand the process. It's reducing some of the work required to turn that understanding into a working system.
For processes that extend beyond a native HubSpot workflow, AI and workflow automation can become part of the larger architecture.
One of the simpler Fall Spotlight updates could also be one of the more practical.
Scheduled Prompts let teams set Breeze Assistant prompts to run automatically on a daily, weekly or monthly schedule. That changes the interaction from “remember to ask AI to do this” to “have AI do this work on a recurring basis.”
Think about the work your team repeats on a schedule: reviewing pipeline activity, summarizing account changes, preparing a weekly sales update, identifying records that need attention or gathering information for a recurring meeting.
Not every recurring task needs a custom agent or complex automation. Sometimes the better solution may be giving Breeze the right context and asking it to perform the same useful work at the right time.
A salesperson finishes a customer meeting and needs to update the CRM, add a note, create a follow-up task and maybe change something on the record. But they're moving to the next thing, so the update gets delayed or skipped.
HubSpot's Breeze Assistant Voice Mode brings CRM-aware AI into that mobile workflow, allowing users to ask about CRM data, update records, create follow-up work and complete other actions using natural language.
If voice shortens the gap between something happening with a customer and that information making it into the CRM, it could improve the data that reporting, automation and AI rely on afterward.
Breeze Artifacts help bridge the gap between asking AI to do something and turning the result into something the business can actually use.
The interesting part isn't that AI can create a document. It's reducing the distance between the conversation and a usable business output.
After everything announced at Fall Spotlight, resist the urge to start with a list of features to turn on. Start with the outcome.
What are you trying to improve?
Then ask whether Breeze, an agent, automation, an integration, a better process or some combination of them could improve that outcome.
For the bigger picture, read HubSpot Fall Spotlight 2026: The Biggest Updates and What They Mean for Your Business.
As AI moves closer to actually doing work inside the CRM, your HubSpot foundation matters more. An agent working with duplicate records still has duplicate records. AI using outdated information still has outdated information. Automation built around a bad process can simply execute the bad process faster.
Before giving AI more responsibility, companies should be thinking about CRM architecture, data quality, integrations, permissions, business definitions, governance and process. Our HubSpot solutions work often starts here.
We've worked in HubSpot since 2013 and across 583 HubSpot portals. Buying or enabling technology is rarely the hard part. Making the technology fit the business is.
That's the approach behind our broader AI solutions work as well. We start with the business requirement, then determine what combination of HubSpot, AI, automation, integrations or custom development makes sense.
Sometimes the answer will be Breeze. Sometimes it will be a workflow. Sometimes it will be an agent or an integration. And sometimes the right answer will be fixing the process before adding any more technology.
Start with the outcome. Then choose the technology.
Tell us what you're trying to improve or what your team is spending too much time doing manually. We'll help you evaluate whether Breeze, agents, automation, integrations or another approach makes sense, and what your HubSpot environment may need before you implement it.
Talk to Primo About AI + HubSpot
Breeze Assistant is HubSpot's conversational AI assistant, designed to work with CRM and business context to help users find information, create outputs and complete tasks across HubSpot.
Breeze Projects group related conversations and allow teams to provide shared instructions and reference material so Breeze can maintain more consistent context across related work.
Agent Builder allows teams to create and customize AI agents for specific business needs. Agent Builder is currently in public beta. Availability and required subscriptions or permissions can vary by capability, so confirm eligibility in your HubSpot portal before planning an implementation.
No. Availability varies by capability, subscription and beta enrollment. Confirm the specific capability and eligibility before planning an implementation.