AI-native product leader and strategic operator · New York

I take products from idea to scale. And I still build the AI myself.

I'm a product leader with 15 years in healthcare, insurance, payments and pharma. I've run an $800M portfolio and built a 40-person product team. Right now I'm designing and building an AI health product myself.

I care about the whole job: the strategy, the customers, the team and the code.

See the work Get in touch

Short on time? The four numbers below say most of it.

From the boardroom to the codebase

Strategy and P&LAn $800M portfolio of 400+ initiatives at New York Life
Leading teams40 people across product, engineering, design and analytics
Pricing and go-to-marketUsage-based pricing for a payments product, then a direct-to-consumer plan for a health app
Customer discovery30+ patient and provider conversations in six weeks
AI architectureSeveral AI models working together, with retrieval on GCP, on FHIR health data
Prototypes in codeFive working AI apps I built myself

$300M+

in added revenue and savings at New York Life, while operating revenue grew from $3B to $3.5B

$0 → $15M

in ARR from a payments product in 9 months, live in 4 countries

30% → 90%

of 4,000+ people using a planning platform I launched

40 people

on the product team I built from scratch, with 90% retention over 9 years

Product, AI and leadership in one person

Most people have two of these. I've spent 15 years building all three, and each one makes the other two better.

Product

Strategy through launch

I own the whole product, not just the roadmap.

Discovery. 30+ interviews shaped the health product. Both New York Life platforms started from user research.
0 to 1. A planning platform MVP in six weeks. A payments product at $15M ARR in nine months. A health MVP in ten weeks.
Pricing and go-to-market. Tiered usage-based pricing, subscription models and a direct-to-consumer launch.
Metrics. OKRs and North Star metrics across 10+ business units, and executive decisions made 5x faster.

Hands-on, not from above

I design the systems and write enough code to prove them.

Architecture. Several AI models (OpenAI, Gemini, Claude) working together, with retrieval on GCP.
Data. Pipelines from 5+ clinical sources into one record on FHIR, with a HIPAA and SOC 2 roadmap.
Building. Five AI apps prototyped myself in Cursor, Replit and Lovable, plus a voice interface with ElevenLabs.
Judgment. MIT Sloan AI strategy and two AI product leadership certifications, plus a habit of asking if users would regret it.

Leadership

Teams and executives

I build the people and the operating rhythm behind the product.

Teams. Built a 40-person product team with 90% retention over nine years, and promoted 7 PMs.
Executives. Chosen by a COO to build a product function, and recruited by a CEO to run a financial products line.
Money. Ran an $800M portfolio and moved $300M toward customer-facing growth.
Operating model. Stage gates, investment reviews, OKRs and QBRs that made speed to market 30% faster.

Six products, from first idea to scale

Each card names the skills it took. Open one for the full story. I've marked what has shipped and what's still a goal.

Adventa Healthtech · 2024 to now · Head of Technology Platform and Product Development MVP shipped, more in progress

Building an AI health product, from first interview to working MVP

I ran the interviews, designed the architecture, set the pricing and go-to-market, and wrote the compliance roadmap. Patients can run it by voice.

DiscoveryAI architecturePricingGo-to-marketCompliance

30+conversations5+data sources10 wksto MVP

The full story

Problem: People fall off their care plans partly because their health information is scattered across five or more systems. After 30+ conversations with patients and providers in six weeks, I stopped calling it a compliance problem. It's an information problem.
What I did: I designed the architecture: several AI models (OpenAI, Gemini and Claude) working together, with retrieval running on GCP. It brings medical records, visit transcripts and readings from devices like glucose monitors and blood pressure cuffs into one shareable health history, built on FHIR and Apple Health. You can run it by voice, so logging something takes almost no effort. I also set the go-to-market, the pricing approach and the HIPAA and SOC 2 roadmap.
What happened: In 10 weeks we went from idea to a working MVP that unifies a patient's data. Insurance claims and the pharmacy refill feature, including an AI that phones the pharmacy for you, are still being built.

Where we're aiming, not what we've hit

92% of patients following their care plans, and 47 to 55% less admin work for providers. The platform is in beta, so treat these as goals.

Design lensPeople act when they want to, can, and get a nudge at the right moment. Voice and one shared record make it easier to "can."

New York Life · 2015 to 2024 Chosen by the COO

The $600M nobody could see

New York Life couldn't see what its biggest spending bets were returning. I built the product function and the platform that could.

Strategy0 to 1Executive alignmentAnalytics

6 wksto MVP30→90%adoption4,000+users

Read the story

Problem: New York Life was running 400+ initiatives worth $800M with no single way to plan or prioritize them. The COO asked me to build the product function and this platform.
What I did: I shipped a first version in six weeks and kept talking to users until adoption took off. A second platform pulled in 100+ data sources, added GenAI summaries and tied everything to company-wide OKRs.
What happened: It found $600M+ in spending nobody could show results for, and helped move $300M toward customer-facing growth.

iQor · 2009 to 2010 · Director of Product, Financial Solutions Recruited by the CEO

A payments business from nothing, in nine months

iQor had no payments product. Nine months later it had $15M in ARR across four countries.

Market discoveryPricingPartnershipsProduct-led growth

$15MARR100+clients in 90 days~90%retention

Read the story

Problem: iQor, a $16B outsourcing company, had no payments product. The CEO brought me in to run all of its financial products.
What I did: I found the gap in the market, wrote the spec, negotiated bank connections in the US, Canada, the UK and India, and set usage-based pricing. Customers could sign up and get started on their own.
What happened: 20,000 transactions a month and $15M in ARR within nine months. It also opened the door to three more products.

iQor · 2009 to 2010 Turnaround

The product that kept beating its owners

Several leaders had tried to fix it. In four to five months, defects fell from 80% to under 10%.

TurnaroundTeam leadershipAgile

3 → 8NPS80 → <10%defects3xmore quality releases

Read the story

Problem: I took over a product with 60% customer churn and an 80% defect rate, while launching payments at the same time.
What I did: I moved the team to short build cycles, ship a basic version first, and set up weekly customer feedback tied to churn and uptime.
What happened: In four to five months, defects dropped below 10%, churn roughly halved and uptime reached 99.9%.

Novartis · 2010 to 2013 Regulated

Moving drug-regulated apps to the cloud with zero violations

Regulators audit these systems closely, and plenty of companies have stumbled moving them. We saved $4.5M a year and stayed clean.

Regulated deliveryGlobal team buildingP&L

$4.5Msaved a year0violations10 → 75ops team

Read the story

Problem: Clinical applications that regulators audit closely (GxP, 21 CFR Part 11) were running on aging infrastructure. Moving them to the cloud is risky.
What I did: I led the migration and built a 24/7 support team across the US, Switzerland and India, growing it from 10 people to 75 in six months.
What happened: Savings of $4.5M a year, deployments 80% faster, no compliance violations and 50% fewer service disruptions.

Independent consulting · 2013 to 2015 Advisory

From 70% quality to 95% in six months

A retailer in Asia-Pacific was losing revenue to quality problems. Research and small experiments found where.

User researchExperimentationFounder advice

70 → 95%quality$12Mrevenue$20Msavings

Read the story

Problem: A retail client had quality at 70% and a customer satisfaction score (NPS) of 6, and it was costing them revenue.
What I did: I ran user research and small experiments to find where things broke, then fixed them one by one. On the side, I advised founders on pricing and positioning, including work that supported a $1M raise.
What happened: Quality reached 95% in six months, with $12M in extra revenue and $20M in savings.

The part most product leaders skip: I build

I'd rather show you something that works than describe it. These five apps tested the idea behind my current platform, and each has an AI assistant built in.

01
Medical billing decoderReads clinical notes to catch billing and coding errors.
02
Health AI assistantTakes in medical records, logs vitals and tracks care plans.
03
Voice-first medication trackerAsk about drug interactions, track refills and get reminders, all by voice.
04
Home inspection risk analyzerScores a property's risks with AI.
05
College visit plannerPlans campus visits, even on short notice.

Stack · Cursor · Claude · Lovable · Supabase · Pinecone RAG · ElevenLabs

The one that taught me the most

I launched it. People didn't stick.

The medication tracker, an early prototype separate from the production platform.

I launched a text version50+ beta users.
People weren't sticking with itI interviewed them. Too many steps, too much to read, and many users were older.
I rebuilt it around voiceTyping became talking, using ElevenLabs.
Activation went up 70%Voice became the core of the platform.

Activation, text version = 100

Text100

Voice170

How I work

A few habits that have held up, from a $3.5B insurer to a pre-seed startup.

Start with the real problem

Before I ask what to build, I ask what's going wrong for the customer and the business. The best answer is often a rethink, not a new feature.

Talk to people first

Thirty conversations in six weeks changed how I framed the whole health product. I do the research before the architecture.

Make the right thing easy

People do things when they want to, they can, and something nudges them at the right moment. The missing piece is usually "can." That's why the medication tracker went voice-first.

Build it, then lead it

I can prototype in code, run a customer interview and present to a board. I've also run a 40-person product team and promoted 7 PMs.

Ask if people would regret it

With health data, I imagine the user in the room. If they'd be uncomfortable knowing how and why it was built, I change the product.

Call a goal a goal

I keep measured results and targets separate, everywhere, including this page.

Where I've worked

Mostly regulated industries, where the stakes are real.

2024 to now

Adventa Healthtech

Head of Technology Platform and Product Development. I own strategy, roadmap, go-to-market and vendor choices.

2015 to 2024

New York Life

Senior Director, Head of Products and Platforms. Ran an $800M portfolio of 400+ initiatives and built two platforms from scratch. Won the company's Presidential Award.

2013 to 2015

Independent consulting

Product and technology advice for insurance, healthcare and retail clients.

2010 to 2013

Novartis

Senior Director, Product Portfolio Operations and Strategy. Ran regulated apps worldwide with a $2M P&L and a 20-person team.

2009 to 2010

iQor

Director of Product, Financial Solutions. Owned the financial products line: $7M budget, 20 people, four bank partners.

2002 to 2009

Alliance Bernstein, BNY Mellon, PwC

Principal Consultant and AVP. Helped large financial services clients modernize their technology.

I take products from idea to scale. And I still build the AI myself.

Product, AI and leadership in one person

Strategy through launch

Hands-on, not from above

Teams and executives

Six products, from first idea to scale

Building an AI health product, from first interview to working MVP

The $600M nobody could see

A payments business from nothing, in nine months

The product that kept beating its owners

Moving drug-regulated apps to the cloud with zero violations

From 70% quality to 95% in six months

The part most product leaders skip: I build

I launched it. People didn't stick.

How I work

Start with the real problem

Talk to people first

Make the right thing easy

Build it, then lead it

Ask if people would regret it

Call a goal a goal

Where I've worked

If the problem is hard and the product matters, let's talk.