For healthcare operating partners

Know which AI bets in the portfolio are worth funding.

A ranked, written view of every AI initiative across your portfolio companies: what to scale, what to fix, what to defer, what to stop, what implementation will actually cost, and who owns the result.

Book a call Test me first, free

Don't take my word for any of it. No board deck, no NDA, no data room. Write me a paragraph on each of the three biggest AI bets in the portfolio. In five business days you get the three weakest assumptions in them, in writing, at no cost. Read it, then decide whether we should talk.

Arvita Tripati

The gates I have been through

Delivered inside their regulated programs
Johnson & JohnsonGileadModernaAbbVieAmerican Red Cross

Their clinical trials, run from inside the technology vendor that delivered them.

Cleared their review, and bought
NHSU.S. Department of Veterans Affairs

An FDA-regulated wearable, through two of the most demanding enterprise buyers there are. Security, privacy, clinical, and procurement all had a veto. None of them used it.

Four functions hold a veto over AI in regulated healthcare: regulatory, quality, security, and privacy. I have held accountability in all four. More than 30 FDA-regulated products through those gates as an operator, not an advisor: AliveCor, LabCorp, Vineti, Korio, Clip Health, endpoint Clinical.

Four gates that stopped being bottlenecks

Each of these is a mechanism, not a number on its own. That is what makes them survive the follow-up question. None of them was solved by a better model.

6 weeks→a day and a half
Client delivery documentation at a healthtech SaaS company, drafted by AI agents. Then the review threshold came down, twice. Project managers stopped writing documents and started carrying more engagements.
4 months→3 weeks
Cut delivery cycle time for a GxP-regulated SaaS platform, through risk-based testing and automation aligned to the release cadence. The client launched early, and put the value of that early launch at $81M.
reactive tickets→7% off support spend
Complaints analysis at an FDA-regulated wearables company that fixed root causes instead of processing tickets.
5 people→17, zero turnover
Built compliance and information security from scratch at a healthtech SaaS company. Audit requests fell 35%. The function returned 57% on its cost.
The operating partner's problem

AI activity is rising across the portfolio. Measurable value often is not.

Companies are launching pilots, signing vendors, adding AI to product roadmaps, and presenting increasingly ambitious plans to their boards. Most still cannot answer the questions the investment committee will ask.

Without those answers, AI becomes a growing collection of tools, experiments, and claims rather than a managed value-creation program. Spend rises. Pilots stall. Vendor costs stack up. The board sees activity, not evidence.

And the weakest assumption in an AI plan is almost never about the AI. It is a decision that crosses regulatory, quality, security, and finance, and belongs to none of them.

What changes after 10 business days

You walk into the next portfolio review with:

Not another AI strategy deck. A set of decisions the operating team can act on.

Six weeks of documentation became a day and a half. The capacity showed up in the P&L.

What it looks like when the work actually leaves the model.

At a healthtech SaaS company, every client engagement opened with the same three documents: a project plan, a requirements specification, and a risk assessment. Project management owned them. It took roughly six weeks to produce them, and the engagement did not move until they were done.

Everything those documents needed already existed. It was in the proposal, and it was in the intake form the client filled out with us at kickoff. The six weeks were not spent finding information. They were spent transcribing it, formatting it, and routing it.

We built agents to draft all three from the proposal and the intake form. The draft came back in a day and a half.

Then the decision that actually mattered. I set the review threshold: what a human still had to read, what they could sign without reading, and what nobody needed to look at again. I made that call, which meant I owned it if it went wrong.

It did not go wrong. At six months we cut review further and put monitoring behind it. At twelve months we cut it again.

Project managers stopped producing documents and started running more engagements. The same team carried more projects.

Most AI rollouts stop at the draft, because the draft is the easy part and nobody has to own it. The work only leaves the model when someone senior enough decides what a human no longer has to check, and then schedules the next reduction instead of waiting to be asked.

Why not the alternatives

  • A strategy firm. You get a deck, a team of associates, and a six-month engagement. You still have to decide. I do the diagnosis myself, in ten days, and end with decisions rather than options.
  • An AI platform vendor. Their diagnosis always concludes that you need their platform. Mine has no product at the end of it.
  • Your operating team. They could do this. They are also running the companies. The reason it has not happened is not capability. It is calendar.

What I do not do

  • Technical AI implementation or data engineering execution
  • EHR integration delivery
  • Generic AI training or policy templates without operating decisions
  • One-time innovation workshops with no follow-through
  • Rubber-stamping an AI story leadership has already decided to tell

An operator, not an AI strategist

AI value creation in healthcare sits where product, workflow, regulation, evidence, data, commercial strategy, privacy, quality, implementation, and organizational design meet. The problems blocking value rarely belong to one function, and adding another tool almost never fixes them.

I have spent nearly two decades inside those functions, bringing more than 30 regulated healthcare products to market.

I have never been a banker, and that is the point. The people who can read a model are not usually the people who have shipped a regulated AI product, sat through the FDA inspection, and watched an enterprise deal stall in a security review. I have. When I tell you an assumption in a value-creation plan will not survive contact with a health system, it is because I have been on the other side of that conversation.

I am not an implementation shop looking for a transformation program. I am not here to validate an AI story leadership has already decided to tell.

My job is to tell you where the value is real, where the assumptions are weak, what to fund, fix, defer, or stop, and what operating model gets the result.

They hired me to find what nobody else would.

“Arvita crushes challenging problems and creates order where there is uncertainty. I hired her, and she became an integral member of the leadership team. I hired her again to join me on the founding team at Korio.”

Chuck Harris
Former CEO, endpoint Clinical (acquired by LabCorp)
COO & Co-founder, Korio

“I love Arvita's discipline and the structure of her thinking. Her work was thorough, well-organized, and genuinely useful.”

Drew Bennett
Innovation Partnerships
University of Michigan

Before any work starts, you see the plan.

A 30-minute callWhere AI shows up in your value-creation plans, which companies worry you, what the board is asking. You ask anything.
A written scope, prepared by meCompanies in scope, what I review, what you receive, the delivery date, the fee. You read it on your own time.
You decideIf it fits, we book a start date. If it does not, you keep the scope and the reasoning.
Book a 30-minute call

Ways to work with me

Fixed scope, fixed date, both agreed in writing before any work begins. No hourly billing, no staffing pyramid, no change orders.

AI Plan Teardown
Back in 5 business daysI do four of these a month. Each one is real work.

No board deck. No NDA. No data. Write me a paragraph on each of the three biggest AI bets in the portfolio: what each is meant to do, what it has cost, and roughly what share of its output still gets human review. That last number is usually the whole answer.

You get back the three weakest assumptions in what you sent, and what I would do about each. Yours to keep, whether or not you hire me. Read it, then decide whether we should talk.

Request the teardown
Portfolio AI Scan
10 business days3 to 10 portfolio companies

For operating partners deciding where to focus across the portfolio.

  • Ranked map of AI initiatives across 3 to 10 portfolio companies
  • The weakest assumption in each company's AI plan
  • The operating constraints most likely to block value
  • Where to put your time, your capital, and specialist support first
The decision this enables: where does the operating team intervene first?
Book a call
AI Operating Blueprint
2 weeks to diagnosis30 days to blueprint

For one portfolio company with real AI ambition, active pilots, or an AI-enabled thesis.

  • Ranked value-creation opportunities
  • Implementation economics
  • Workflow and adoption constraints
  • Product, data, evidence, and claims requirements
  • Decision rights and named owners
  • Board-level measures and a 90-day execution plan
The decision this enables: what does management scale, fix, defer, or stop?
Book a call
Fractional Chief AI Officer
6 to 12 monthsI make the decisions, not just recommend them

The title is the shorthand. The substance is decision rights. Every other engagement on this page ends in a document that somebody else still has to sign. This one does not. You give me authority over a defined set of calls, and I make them.

Which calls are mine is the scope, and we agree it before I start. The list below is where those decisions usually sit. We cut it, add to it, and set the escalation path. Then it is signed, and the accountability is real.
  • The review threshold. What a human still has to touch, and when that falls again
  • What gets funded, what gets deferred, and what stops
  • The evidence, claims, and controls a regulated product has to clear
  • The escalation path when regulatory, quality, security, or clinical says no
  • Board and IC reporting on what the AI is actually doing to the business
The decision this enables: the one nobody will make, because their name goes on it.
Book a call
AI Value-Creation Advisory
Monthly retainertypically 6 to 12 months

For partners, CEOs, and boards that need senior judgment while the company executes.

Monthly executive decision sessions, initiative reviews, business-case pressure testing, vendor and partnership evaluation, roadmap reprioritization, adoption reviews, board and IC preparation, value-realization measurement, and early warning on product, regulatory, or operating risk.

Book a call

Frequently asked

A strategy engagement ends with options. This ends with a ranked set of decisions, named owners, and a 90-day plan. If you want a deck, hire a strategy firm.

Board decks, operating data, product plans, and 60 to 90 minutes each with a handful of executives. I work from what already exists. I do not run weeks of workshops.

Then you have that in writing, with the evidence, before your next investment committee. A confirmed plan is a result, not a consolation prize.

AI should not be judged by how many pilots a company launches.

It should be judged by whether it improves the business. If AI is in the value-creation plan but no one can show where the value is coming from, the next step is not another pilot. It is a better operating decision.

Book a call Request a free teardown

Prefer a peer room? The CEO Circles  ·  Looking at a deal rather than a portfolio? Read the diligence page  ·  Running a portfolio company? Read the CEO page