You've experimented with AI. Now decide what it should change.
I help healthcare administration teams and growing businesses assess what they've tried, set a clear vision for where AI goes next, and build it with their team.
Free, 30 minutes. If I can't help, I'll say so on the call.
Healthcare administration
Health plans, TPAs, benefits administrators, and provider services. Pilots that make it through compliance review and a roadmap leadership can defend.
See how I work with healthcare teams
Growing businesses
Firms of 10 to 200 people. Know which AI bets pay off, then make them part of how you work.
See how I work with growing businesses
The AI ladder
Most organizations stall at L2.
People use ChatGPT or Claude every day. A few pilots are running. And how the business actually runs hasn't changed. That gap between individual use and AI built into the work is where most teams stop.
I get you across it and further if you want to keep going.
L0
AI-absent
No AI in use
L1
AI-curious
AI inside tools you already pay for
L2
AI-active
People use ChatGPT or Claude on their own
L3
AI-integrated
Workflows with AI built in, owned and measured
L4
AI-native
AI is core to how you deliver
Explore
Execute
Three questions come before any technology
Every engagement starts with the same three business questions. None of them are technical.
- 1
What is your business trying to hit this year?
- 2
What are the things standing between you and that number?
- 3
How much are those pain points costing you?
Answer those and you know what to build. A few weeks later, you'll know whether it's working. Skip them and you get AI theater: an impressive demo nobody uses.
Measured on outcomes, not activity
Before anything gets built, we agree on the numbers that define success and record where they stand today.
Time: elapsed time from request to done, not just hands-on minutes.
Quality: error and rework rates.
Touches: how many hands and handoffs one piece of work needs.
Cost: cost per unit of work, tracked from day one.
Adoption: how much of each AI draft survives review. That number should climb every week.
Every build commits to an outcome: at least one workflow in production, measured against that baseline. If there's room to do more inside the agreed scope, we do more.
How it works
I look at the work before I change any of it.
01
01Assess
What you've tried, what's working, what's stalled, and what to stop.
Your part
An hour with leadership, plus access to the pilots.
02
02Envision
Where AI should take the business in the next 12 to 18 months and which workflows get you there first.
Your part
One leadership working session.
03
03Build
The first workflow in everyday use, built with your team so they learn by doing it.
Your part
One named owner, 2 hours a week.
04
04Govern and manage change
Ground rules your team will actually read and a plan that brings the rest of the team along.
Your part
Sign-off and a hand in building it out.
05
05Expand(optional)
The next workflows, a refreshed roadmap, and ground rules that grow with you.
Your part
A standing check-in, as often as it's useful.
Services
3 to 4 weeks
AI Vision & Roadmap
Where you are, where to go, what to build first.
About 90 days
Workflow Build
One or two workflows moved into everyday use.
Optional, ongoing
Expansion Partner
For teams that want to keep going.
Half day or less
Leadership sessions
Visioning workshops and talks.
Up to 75%
faster clinical report generation from a GenAI documentation workflow I built and shipped.
$100M+
in medical claims supported at a self-funded health benefits platform, where I was the first product hire.
1 FTE
of capacity added at a marketing agency by rebuilding its core production workflow with AI.
What I don't do
I don't sell platforms.
No software to subscribe to, no vendor commissions. My tool recommendations have no stake in the answer.
I don't replace what works.
If a tool does its job, we connect to it.
I don't write shelf strategy.
Every roadmap comes with a first workflow in production.
I don't hold the keys.
Accounts stay in your name. Removing my access breaks nothing.
Your data
Your accounts, in your name. I only touch the information a job needs. In healthcare, member data stays out of AI tools unless the right agreements, including a BAA, are in place. I'll tell you on the first call exactly what's required.
Louis Levine
I've led strategy teams at Deloitte and CVS Health and built products at several health tech startups, two of them acquired. Now I help organizations turn AI experiments into how the work gets done.
Tell me what you've tried so far.
Thirty minutes, free. I'll tell you honestly what I'd do next, even if it isn't hiring me.