What an AI enrichment agent does before every HubSpot sales call

Most sales teams don't have a lead problem. They have a preparation problem. The contacts are already in HubSpot, but each one arrives as a name, an email address and not much else, and a rep has a few seconds to decide whether it's worth a call and what to say if someone picks up.

An enrichment agent closes that gap. It's the first of the three AI agents I built into Dale Carnegie's HubSpot, and its whole job fits in one line: it preps every contact before the call.

What problem does an enrichment agent solve?

It takes contact research off your reps. When I started with Dale Carnegie, 7,000 contacts sat in HubSpot with plenty of volume and no working priority system. Reps were guessing who to call next. Every minute spent looking someone up, or scrolling back through old notes to remember the last conversation, was a minute not spent on the phone.

That's manual CRM data entry in its most expensive form: skilled salespeople doing research and admin. It's also the easiest work to hand to an AI agent, because it's repetitive, it follows the same pattern for every contact, and it has to happen before every single call.

What does an AI enrichment agent do in HubSpot?

It works on the contact record itself, ahead of the rep. The specifics depend on your funnel, but the work usually covers:

• Filling the gaps in the record, so the fields your pipeline and lead scoring depend on aren't left blank.
• Bringing the contact's history into one place: past calls, emails and deal activity, summarized instead of scattered across the timeline.
• Flagging what the rep should know before dialing, such as whether the contact has been called before and how that went.
• Keeping the record clean as it goes, so the next agent and the next report work from the same data.

The point isn't a longer record. It's a rep who opens the contact and knows, in a few seconds, who this is and why they're calling today.

When does the enrichment agent run?

Continuously, in the background, just ahead of the rep. It isn't a one-off data cleanup, because enrichment only pays off if it happens at the right moment. At Dale Carnegie it works alongside a call priority queue in HubSpot: the queue decides who gets called next, and the enrichment agent makes sure that contact is ready before the rep gets there.

It's also the first link in a chain. The coaching agent reviews the call once it happens, and the re-engagement agent follows up with anyone who doesn't book. Each agent works better because the one before it did its job.

What changed at Dale Carnegie

Between February and July 2026, inside the same HubSpot instance and with no added headcount:

• Contacts under active management grew from 1,703 to 11,413 (6.7×).
• Calls made per month went from 633 to 1,247, nearly double.
• Meetings booked per month went from 5 to 91 (18.2×).

Enrichment didn't do that alone. Those numbers are the compounding effect of the priority queue and all three agents working together. But call volume doesn't double on the same headcount while reps are still researching every contact by hand, and that's the work the enrichment agent took off their plate. The full case study has the rest of the numbers.

Do you need an enrichment agent?

An enrichment agent is worth building if any of these sound familiar:

• Reps spend time looking contacts up before they call.
• Key fields in your CRM are regularly empty, so lead scoring and reports can't be trusted.
• You have more contacts than your team can research, and the ones they skip are the ones that go cold.

It's rarely the first thing I build, though. Every engagement starts with a diagnosis of where revenue is actually leaking, then the pipeline and reporting foundation, and only then the agents. An enrichment agent on top of a messy CRM just produces messy data faster. If you want to know whether your HubSpot is ready for one, start with a short call.