Most firms in this space will not put a number on a page, which wastes everyone's time. Here is ours, what is inside it, what is not, what grants can take off it, and how to think about whether it pays back. If the number is wrong for your agency, you will know in about ninety seconds instead of after three calls. For what the money actually buys, in sequence, read our practical guide to AI for insurance agencies.
What does AI training for an insurance agency cost?
The Foundation starts at $12,000 for the first 10 seats, plus $500 per additional seat, and that includes the first 60 days of the Integrated AI Partner retainer. After that the retainer is $1,500 a month for up to 10 people, $2,000 for 11 to 25, and $2,500 for 26 and up, month to month with 30 days' notice. State incumbent-worker training grants can often reimburse 50 to 90 percent of eligible training cost when applied for before training starts. Pricing is valid for engagements signed through December 31, 2026.
The Foundation is priced as a piece of work with a seat count, not a per-seat license. Ten seats are in the base price because the effort sits in the audit, the tools and rules, and building the workflows, not in how many chairs are in the room. Past ten, each seat adds $500, because each person is another set of accounts to audit, another set of tools to configure, and another chair on the two on-site days. An 18-person agency, for example, is $16,000 for the Foundation and then $2,000 a month from month three.
What is included in the Foundation fee?
A data and exposure audit of every tool and account in use, approved tools, PHI-safe configurations and an acceptable-use policy in place, your first 5 to 10 workflows built on your real accounts, two days on site with the whole team, a shared prompt and skill library the agency owns, and the first 60 days of the Integrated AI Partner retainer.
- Discovery and the audit. Which workflows would move first, every tool in use, every account, what data has gone where, and what it is costing you today. Week one.
- The platform, tools and rules. One recommendation on which platform to consolidate on, then approved tools configured, PHI-safe settings, and an acceptable-use policy written so the safe path is also the fast one. Week two.
- The build. 5 to 10 workflows built in your own AI environment and loaded into a library the agency owns. This is the largest share of the effort and the part that makes the on-site days land.
- The two days on site. Your whole team, your systems, your live accounts. Week three.
- What you keep, and what keeps going. The tools, the policy, the library, the workflows, Hub access, and 60 days of the retainer: two monthly briefings before the first invoice.
The full engagement is described on the Foundation page, and what happens after it is covered in training is the foundation.
What is not included?
Two things. Your AI tool subscriptions, which you buy directly and own, and your team's time. Budget a few hours in week one for the audit with the people doing the work, plus the two on-site days themselves.
The time cost is the one that gets underestimated, so it is worth saying plainly. The audit only works if we get the account managers and producers who actually run the work, not a manager describing it secondhand. And the on-site days need to be real days. An agency that books them and then lets people drift in and out of the room gets a worse result, and we would rather move the date than run it that way.
Can grants cover part of it?
Sometimes. State incumbent-worker training grant programs are designed to reimburse employers for training existing staff, commonly at 50 to 90 percent of eligible cost. The Foundation is part audit and set-up and part training, so what qualifies is the training portion, and the sequence is unforgiving: they generally must be applied for before training starts.
That rule decides how much of the $12,000 is really yours to carry, so it is worth resolving early rather than late. Our guide to state training grants covers how the programs work, what they usually ask for, and how to find your state's version. On the call we will flag whether it looks applicable in your state, though the application itself is yours to run with your state agency.
How quickly does it pay back?
Run it on your own numbers: groups renewing, times hours per renewal, times loaded hourly cost, is what one workflow costs you today. An agency with 60 groups at 20 hours each spends 1,200 hours a year on renewals. Recover a third of that and it is 400 hours back in year one, which at a loaded cost of $40 an hour is $16,000. In one documented engagement, a five-person benefits team recovered 7 to 13 hours per week, roughly $32.5K in annualized capacity conservatively, with Phase 1 payback in under 9 weeks.
Two honest caveats. The documented number is a single engagement, so treat it as an existence proof rather than an average. And recovered hours only turn into money if they go somewhere: more clients served, work that used to get skipped, or capacity you would otherwise have hired for. An agency that recovers ten hours a week and lets them evaporate has bought nothing.
The arithmetic that matters is your own, and the benefits page has a calculator for it. We publish the before-and-after times we have measured in the five use cases every agency builds in week one so you can run that math before you talk to us. Whatever you land on is for one workflow. The Foundation builds 5 to 10.
What does it cost to keep going?
The Integrated AI Partner retainer, which works like your own AI department, is $1,500 a month for up to 10 people, $2,000 for 11 to 25, and $2,500 for 26 and up. It covers monthly briefings, quarterly refreshers, on-call support, and new workflows as your needs grow. The first 60 days are included in the Foundation, and after that it is month to month with 30 days' notice. If you stop, everything built during the Foundation stays yours.
Most agencies decide this after the two included briefings rather than before, which is the right time. The details are on the Integrated AI Partner page.
