Breeze · HubSpot AI

HubSpot’s AI, working on your real data

Breeze Assistant, the agents in Agent Hub and the AI built into every Hub are only as good as the CRM they read from. We tidy up your data, decide what the AI should do for each team and roll it out under control, with no nasty surprises on your credits bill.

HubSpot Diamond Solutions Partner
4.9 Verified reviews on HubSpot
9 official HubSpot
certifications
+200 clients in Spain,
Italy and Dubai
19 people on the team
15 in Barcelona · 4 in Dubai

Before you switch on AI

AI won’t tidy up your CRM. It amplifies what’s already there.

Turning on Breeze takes a few minutes. Getting useful answers out of it is another matter: an agent reading duplicate records, empty properties or deal stages every rep uses differently will give you answers that are just as messy, only faster.

That’s why we start with what the AI is going to read. We review data quality, the properties people actually use, your sales and service processes, and the material that should feed the agents: knowledge base, FAQs, policies and brand voice.

With that foundation in place, AI stops being a flashy demo and starts saving your teams real hours.

What AI needs to work well

  • Contacts, companies and deals without duplicates
  • Clear properties, filled in consistently
  • Sales and service processes written down
  • Up-to-date help content in your languages
  • Owners who review and measure what each agent does

HubSpot AI today

Breeze, agents and credits, minus the marketing

HubSpot reshuffled its naming in 2026: agents are now managed from Agent Hub, and Breeze Studio became Agent Builder. What’s available depends on your edition, so we tell you what you already have before recommending anything.

The assistant: Breeze Assistant

HubSpot’s built-in conversational assistant. It summarises records, preps meetings, drafts content and answers questions using your CRM data, within each user’s permissions.

HubSpot agents in Agent Hub

Ready-made agents such as the Customer Agent, the Prospecting Agent and the Data Agent. Some of them require a Professional or Enterprise edition.

Custom agents with Agent Builder

For use cases specific to your business: agents with their own instructions, knowledge sources and actions, plus agentic workflows on Professional and Enterprise editions.

AI built into every Hub

Features inside the tools you already use: call and email summaries, suggested next steps, and help drafting emails, landing pages or ticket replies.

HubSpot Credits

Many agents and AI features consume credits. Depending on your subscription there’s an included allowance that can be topped up; we set limits so usage stays predictable.

AI settings and permissions

In the AI settings, admins decide which features are switched on, for whom and which data they can access. It’s the foundation of AI governance in your portal.

Use cases by team

Where AI adds value across your business

We don’t switch everything on at once. We pick two or three use cases with a clear payoff, measure them and expand from there.

Sales

  • Meeting prep with the full account history
  • Call summaries and next steps logged on the deal
  • Prospecting based on buying signals, with tailored messages
  • Human review before anything is sent

Marketing

  • Drafts of emails, landing pages and posts in your voice
  • Campaign ideas built on your real segments
  • Content repurposed across formats
  • Editorial sign-off before publishing

Customer service

  • Customer Agent on chat, email or other channels
  • Answers grounded in your knowledge base
  • Hand-off to a person with full context
  • Ticket summaries and suggested replies

Data quality

  • Properties filled in from calls and emails
  • Incomplete or duplicate records flagged
  • Plain-language questions to the CRM
  • Rules the AI can’t override

Operations and RevOps

  • Workflows with AI steps where they help
  • Request classification and routing
  • Custom agents for repetitive tasks
  • Integrations that give agents context

Leadership

  • Questions about pipeline and customers without waiting for a report
  • Key account summaries before a meeting
  • Credit usage tracking
  • A dashboard showing the impact of AI

Governance and GDPR

AI with clear rules from day one

An agent answering your customers speaks on behalf of your company. Before switching it on, we agree with you what it can do on its own, what needs a person to sign off and what information it must never touch.

HubSpot lets admins control which AI features are used and what data is shared from the AI settings, and its Trust Center explains how data is handled in its AI features. We turn that into rules that fit your business: permissions by team, sensitive data kept out of agents’ reach, transparency with customers when they’re talking to an AI, in line with the EU AI Act, and a review with your data protection officer.

What we put in writing

  • Which AI features are enabled, and for which users
  • Which data and sources each agent can access
  • When an agent must hand over to a person
  • A clear notice to customers when an AI is helping them
  • Credit usage limits and regular reviews

Measuring impact

If you don’t measure it, it’s just a demo

We take a baseline before switching on each use case and compare afterwards. That way you know which agent earns its credits and which one needs tuning or turning off.

Customer service

What the agent resolves and what reaches your team.

Conversations resolved without helpFirst response timeHand-offs to a personCustomer satisfaction

Sales

Selling time recovered and opportunities created.

Meetings bookedReply ratePrep time per accountDeals with a next step

Data

A more complete, reliable CRM month after month.

Key properties filled inDuplicates flaggedRecords with no ownerData updated after every call

Cost and adoption

What AI consumes versus what it saves.

Credits per use caseEstimated hours savedActive assistant usersAnswers corrected by the team

How we set it up

From curiosity to AI in production in four phases

If your portal is already well implemented, we go straight to the use cases. If not, we sort out the CRM first: without that foundation, no agent performs well.

Audit

AI readiness

We review data, properties, processes, help content and your HubSpot edition, then tell you what you can switch on now and what needs fixing first.

Data quality reportPrioritised use cases

Foundation

Data and knowledge

We clean and standardise records, tidy up properties and prepare the sources the agents will draw on: knowledge base, policies, tone and business context.

Duplicate-free dataKnowledge sources

Pilot

Setup and testing

We configure the assistant and the chosen agents with permissions, hand-offs and credit limits, and test them on real cases from your team before opening them up to customers.

Permissions and limitsTests on real cases

Scale

Measure and expand

We train each team, measure against the baseline and fine-tune instructions and sources. Then we extend to new use cases or custom agents.

Impact dashboardRole-based training

Why Vexco

Useful AI, not window dressing

01

Data first

We come from large-scale migrations, deduplication and complex data models. We know how to get a CRM into the shape AI needs to answer well.

02

Process before tools

We define what each agent should do within your sales or service process, not the other way round. If a use case doesn’t pay off, we’ll tell you.

03

Consultants and developers together

The same team configures HubSpot and builds integrations with your ERP or other systems, so agents get the context that currently lives outside the CRM.

04

A Diamond Partner with public reviews

We’re a HubSpot Diamond Solutions Partner, with 9 certifications and a 4.9/5 rating in HubSpot’s official partner directory.

Case studies published by HubSpot

Don’t take our word for it. Take HubSpot’s.

HubSpot has published these implementations as official customer stories, with Vexco as the partner. The figures are the ones HubSpot reports.

Lexidy

Legaltech · Barcelona

+75%higher conversion rate

Automatic lead assignment by nationality and budget, process automation, and tickets that track each lawyer’s workload. Today they operate in 6 countries.

  • Sales Hub
  • Marketing Hub
  • Service Hub
  • Data Hub

Winahost

Holiday rental management

−38%fewer manual touchpoints

From spreadsheets to automated sales and customer service: time to close down 12% and the portfolio managed by Customer Success up 46%.

  • Sales Hub
  • Service Hub

FAQs

Questions about Breeze and HubSpot AI

What is HubSpot Breeze?

It’s the name HubSpot launched its AI under: Breeze Assistant, the agents and the AI features built into the Hubs. In 2026 HubSpot reorganised the brand: the assistant is still called Breeze Assistant, while agents are managed from Agent Hub, with Agent Builder for creating your own.

Which HubSpot edition do I need to use the AI?

HubSpot offers Breeze Assistant on all its plans, although some of its features depend on your edition, and several agents and agentic workflows require Professional or Enterprise. We check your portal during the audit and tell you what you already have.

How do HubSpot Credits work?

Many agents and AI features consume HubSpot Credits. Depending on your subscription there’s an included allowance, which can be extended with credit packs or pay-as-you-go. We set up limits and tracking so you know what each use case consumes.

Is it GDPR compliant?

HubSpot gives admins controls over which AI features are used and what data is shared, and explains in its Trust Center how data is handled. Compliance also depends on how you configure it: permissions, sensitive data, what you tell your customers and a review with your data protection officer. We help you get it all defined.

Can the AI answer our customers without supervision?

It can, but we don’t recommend starting that way. We begin with a pilot in which the Customer Agent handles a limited set of topics, hands over to a person when it isn’t sure, and we review its answers before widening its scope.

Do we need to clean up the CRM before switching on AI?

Almost always. Agents read your contacts, companies, deals and help content; if those are duplicated or incomplete, the answers will be too. During the audit we measure data quality and tell you what to sort out first.

How long does it take to get up and running?

It depends on the state of your portal and the use cases. With a well-implemented CRM, a first pilot can go live within a few weeks; if data needs cleaning or processes need rebuilding first, we plan it in phases within the same project.

Free audit

Is your sales process working for you?

In 30 minutes we review how you capture, follow up and close opportunities, in HubSpot, another CRM or Excel, and tell you honestly what we would change first. No commitment, no sales pitch.

We reply within 24 business hours.

Free audit