Breeze AI implementation and management

Put Breeze AI to useful work

We help you put Breeze to work on specific jobs, with a clear purpose, controlled usage and a way to judge the result.

Quick answer

Breeze AI implementation starts with one defined job, suitable data and a person accountable for the result. Project36 checks feasibility, prepares knowledge sources, configures the use case and tests responses or actions with your team. Human review, permissions and running costs remain part of the design.

Key takeaway: Use AI when it improves a real task. A simpler workflow or existing feature may be the better answer.

Is AI the right way to solve this?

Choose a next step based on your situation, including its limitations
Your situationA useful next stepWhat to keep in mind
The task needs help interpreting or drafting informationAssess a focused Breeze use caseOutputs need testing against real examples and a clear escalation path; don't assume every answer is correct.
The task follows consistent rules and needs a predictable outcomeReview the existing process and automationA workflow may be more predictable than AI, but still needs clean data, ownership and testing.
You can't explain the current credit consumption or expected costReview usage before expanding AIAvailability and consumption depend on your HubSpot subscription and the actions used. Check your current account and quote.

AI that does useful work, not demos

The useful question isn’t how many AI features you can switch on. It’s where assistance, automation or better access to information can improve a real task without weakening trust or control.

We prepare data and knowledge sources, configure agents, test responses and actions, set human handovers and train your team. Then we keep improving how they perform.

Guide

AI for marketing teams

Why marketing AI keeps producing generic work, and how to give it your business context.

Read more

What better can look like

  • Prioritised Breeze use cases

  • Less time spent on repetitive work

  • Clear human review points

  • Better prepared data and knowledge

  • Visible usage and running costs

From useful question to working change

Find the useful work

We identify repeated tasks, decision bottlenecks and information gaps worth exploring.

Assess feasibility

We examine data, permissions, platform capability, risk and expected benefit.

Build and test

We configure a focused use case and test it with the people responsible for the outcome.

Enable the team

We document boundaries, review points and the practical way to use it.

AI supporting adoption

For Artisan, behaviour-driven education helped reduce three-month break-clause activations by 65% and increased feature adoption 2.3 times.

Read the client story

Useful preparation

Useful things to know

  • No. A well-defined task can be a sensible place to start, provided the data, risk and expected benefit are understood.

Related ways we can help

  • HubSpot consultancy and optimisation

    We find the friction, underused capability and avoidable admin holding your team back, then turn the best opportunities into practical improvements.

    Find out more
  • Integrations and custom development

    We connect HubSpot to the tools your business depends on and build custom integrations when a suitable connection doesn’t already exist.

    Find out more
  • Practical HubSpot training

    We train people using your setup, your processes and the work they actually do, so learning turns into better everyday habits.

    Find out more

What do you wish HubSpot could do?

Tell us about the time-consuming task, disconnected system or idea you haven’t quite worked out yet.

AI inside HubSpot

AI needs a defined job, appropriate data access and someone accountable for the result. Explore use cases alongside governance and consumption, not in isolation.