Customer Agent
Answers customer questions in chat and email from your knowledge base, and hands off to a person when it can't.
AI Readiness Assessment for HubSpot
HubSpot's agents work from whatever your portal says. We read your real data, show what has to be cleaned up before an agent can run well, tell you which agents fit your business, and put a credit cost on each one before you switch anything on.
Try your numbers
HubSpot bills agents in credits when they complete work. Move the sliders to see a month at your volumes. Your assessment builds this from your real data.
Estimate only. HubSpot rates as of November 2026: 50 credits per resolved Customer Agent conversation, 100 per Prospecting Agent lead, 10 per Data Agent prompt, about $0.01 a credit on pay-as-you-go. Rates change; your assessment checks the current ones.
How it works
With your approval, a read-only review inside HubSpot, plus a few screenshots of settings that can't be read any other way. Nothing changes in your portal.
One meeting on what you want agents to do, who would own them, what must always have a person's approval, and what spend makes sense for a pilot.
Your score, the checklist for each agent, the cost estimate and the plan, presented live and left with you as an interactive report.
The 12-month plan
Every step has a checkpoint, so nothing customer-facing goes live until the portal underneath it is ready.
Why readiness comes first
Your team knows which contacts are duplicates, which list is the real one and which lifecycle field to trust. An agent doesn't. It works from whatever the portal says, at scale, and every gap shows up in what it writes, who it contacts and what it answers.
A portal that has served your team well for years can still need work, because agents ask more of it than people do: clean records, segments built on properties that are filled, one process setting lifecycle stage, a knowledge base that covers real questions, and consent set up for automated outreach.
Most of the work behind a good agent happens before it's switched on.
Illustrative figures. Your assessment measures your own.
What we look at
We read your portal against each one and measure it, so every finding comes with a number you can check.
Are the fields agents read, like owner, lifecycle stage and company, actually filled in?
How many contacts and companies exist twice, so an agent would reach them twice?
Are the segments an agent would work from active and built on reliable properties, or old static lists?
Is there one clear process setting each stage, or several workflows overwriting each other?
Does your knowledge base answer the questions customers ask most, and is it current?
Are subscription types and legal basis set up for the outreach an agent sends?
Which workflows and integrations would collide with what an agent writes?
What you use today, what's included in your subscription, and where AI spend could hide.
What you get
Presented in a live walkthrough and handed over as an interactive report you can share with your team.
Strong knowledge base, but segments and duplicates hold back outbound agents.
Ready to goFix firstNot a fit yet
Illustrative. Uses the calculator's default volumes at about $0.01 a credit.
This is a sample. Yours is built from your own portal.