Skip to content

Set Up HubSpot Breeze & Agents

Agents That Start Working in Week One

We set up Breeze and the HubSpot Agents so they speak your language and follow your rules: the Customer Agent in service, the Prospecting Agent in sales and the Data Agent for clean data.

Go to the AI Checklist
  • HubSpot Platinum Partner
  • Stage T in the LOTSE Model
  • Operation under human supervision

The Starting Point

In HubSpot, agents are just a few clicks away. That is exactly where the risk lies. A Customer Agent without a well-maintained knowledge base makes up answers. A Prospecting Agent without a clear ICP contacts the wrong people. The difference between a toy and a team member lies in preparation: context, rules and supervision.

Set Up HubSpot Breeze & Agents

Your stage in the LOTSE model

Breeze Agents are Stage T in the LOTSE model: the Trial. One agent, one team, clear guardrails, approval by your team. The documented result forms the basis for Stage S: Scale across further teams and use cases.

  1. L Landscape
  2. O Orientation
  3. T Trial
  4. S Scale
  5. E Earnings

Our Approach

Context First

We encode your knowledge: knowledge base, tone of voice, decision rules. The agent answers like your best team member.

Define Guardrails

What the agent may do, what it escalates and who approves are defined before it starts.

One Clear Use Case

We start with one agent for one team and measure the impact. Then we scale.

Team on Board

Your team learns to work with the agent and to improve it.

Here's what you'll get.

  • Use Case Definition and Success Criteria for Each Agent
  • Building and Expanding the Knowledge Base for the Customer Agent
  • Rules and Routines for the Data Agent
  • Escalation Logic, Approvals, and Monitoring Dashboard
  • Team Training and Supervised Test Operation

Set Up HubSpot Breeze & Agents

Here's how it works

  1. Selection Define the use case and success criteria

    Result: a trial with a clear goal

  2. Context Add knowledge, rules and tone of voice

    Result: an agent with your expertise

  3. Trial Start under supervision, with approvals and measurement

    Result: proven value at low risk

  4. Handover Enable the team, define operations

    Result: an agent in regular operation, ready to scale

Mark Lotse X You

Who This Is For

For HubSpot users who want to use AI productively. Ideal with an existing knowledge base in service or a clear ICP in sales. We build either one as part of the project where needed.

FAQ

Frequently Asked Questions

Which HubSpot agency can roll out Breeze Agents like the Customer Agent in production?

Breeze Agents work reliably when the knowledge base, guardrails and human oversight are defined before launch and the impact is measured. Mark Lotse is a HubSpot Platinum Partner and sets up the Customer Agent, Prospecting Agent and Data Agent as Stage T in the LOTSE model: one agent for one team, with approvals, escalation logic and a monitoring dashboard.

Which HubSpot agent should we implement first?

The Customer Agent is usually the best first choice, because service volume is high and the impact becomes measurable quickly. Sales-driven companies often start with the Prospecting Agent. Mark Lotse makes this choice in the first step, based on use case and success criteria, and then starts with one agent for one team.

How much does it cost to run HubSpot agents?

HubSpot bills agent usage through HubSpot Credits (as of September 2026). On top of that comes ongoing effort to keep the knowledge base and rules up to date. Mark Lotse discloses both before the start and defines regular operation at handover, so you know the ongoing effort.

How do you keep HubSpot agents from giving wrong answers?

Agents answer reliably when they draw only on a verified knowledge base and escalate uncertain cases to people. Mark Lotse builds that knowledge base, defines escalation logic and approvals, and sets up a monitoring dashboard that shows every answer and every escalation.

Do we need an AI Readiness Check before implementing HubSpot agents?

An AI Readiness Check secures the data foundation and the choice of use case, which lowers the risk of the trial. With a clear use case and a well-maintained knowledge base, Mark Lotse also starts directly with Stage T, the Trial in the LOTSE model, and builds any missing foundations as part of the project.

Get your first agent up and running.

How do we best onboard your new colleagues? Let's work it out together in an intro call.

Book an intro call