Episode Overview
Podcast: Building One with Tomer Cohen
Episode: Building Heidi With Thomas Kelly: AI As A Care Partner, A Surgeon's Lessons For Building, And The Future Of Healthcare
Date: February 17, 2026
Guests: Dr. Thomas Kelly (Co-founder & CEO, Heidi Health)
Host: Tomer Cohen (Chief Product Officer, LinkedIn)
This episode features Dr. Thomas Kelly, vascular surgeon turned health-tech founder, discussing the journey of building Heidi Health, a leading AI care partner for clinicians. The conversation explores the challenges and breakthroughs in bringing AI to healthcare, doubling clinical capacity, the value of personalization and trust in medical AI, and the future landscape powered by these technologies.
Main Themes
- Translating Clinical Expertise into Product Building
- Pain Points of Healthcare: Administrative Burden
- Building and Scaling Heidi Health
- Product Differentiation & Personalization
- AI Product Go-To-Market in Healthcare
- Impacts on Clinical Outcomes and Healthcare Capacity
- Vision for the Future of Healthcare
- Practical Lessons for Builders
Key Discussion Points & Insights
1. Medicine to Product: The Surgeon’s Perspective
-
Economy of Movement in Surgery and Product (00:06, 06:01)
- Dr. Kelly relates surgical precision to product design:
“There’s this term they say which is called 'economy of movement.' So it’s basically like no wasted movements... I have come to value when there is an elegance or a connection in how things and intention move from one action to the next.”
(Thomas Kelly, 00:06 & 06:01)
- Dr. Kelly relates surgical precision to product design:
-
Motivation for Medicine & Building (02:36)
- Inspired by his GP, admired for knowledge and warmth:
“It always felt like the best use of my talents... would be to do [medicine] for others.”
(Thomas Kelly, 02:36)
- Inspired by his GP, admired for knowledge and warmth:
2. Identifying the Real Problem: Doctors as Paper Pushers
-
Administrative Overload (03:39)
- Dr. Kelly describes the cognitive overload and wasted talent:
“It always felt crazy to me that the doctors were the one doing it. It’s like the pointiest end of the spear... doing very low cognitive effort, very frustrating kind of work.” (Thomas Kelly, 03:39)
- Dr. Kelly describes the cognitive overload and wasted talent:
-
Catalyst for Building Heidi (04:49)
- Personal dissatisfaction with status quo and insight into product needs; initially not motivated by startup culture or VC trends.
3. Building Heidi Health: AI as a Care Partner
-
Core Proposition of Heidi (07:24)
- “Heidi’s an AI care partner for clinicians. The reason we say care partner is there’s actually quite a lot of things that Heidi can do now on behalf of clinicians... we listen into a visit, we transcribe it and generate clinical notes and documents and a Patient summary. We also are starting to do things like calls... helping them schedule in...”
(Thomas Kelly, 07:24)
- “Heidi’s an AI care partner for clinicians. The reason we say care partner is there’s actually quite a lot of things that Heidi can do now on behalf of clinicians... we listen into a visit, we transcribe it and generate clinical notes and documents and a Patient summary. We also are starting to do things like calls... helping them schedule in...”
-
Viral Adoption & Global Footprint (07:24)
- Savings for clinicians: sometimes 1–2 hours per day
- Expansion from Australia to Canada, UK, ANZ, and US
-
Doctors at the Center (08:36)
- “Overwhelmingly clinicians still want to be doctors. They want to do the fun bit, which is figuring out what’s going on... doing the surgery. These are the best parts of the job.”
4. Major Technological Shifts and Product Focus
-
AI Breakthroughs: GPT’s Arrival (10:19)
- Initially a threat, then an opportunity:
“At first it was definitely the, oh shit, this is so scary because all of our technical advantage is gone. But then it became exciting because we could focus on being more of a product centric company.”
(Thomas Kelly, 10:19)
- Initially a threat, then an opportunity:
-
Pivot to Product-Led Growth (10:19)
- The release of powerful AI models made unique tech less defensible, so the team doubled down on personalization, a freemium/pro user template system, and direct adoption by clinicians.
5. Differentiation: Personalization and Precision
-
High Quality, Deep Personalization (12:19)
- “Heidi should be writing within 1% of what you would write reliably, consistently, along with all the other engineering about how you handle offline, low Internet, different languages.”
(Thomas Kelly, 12:19)
- “Heidi should be writing within 1% of what you would write reliably, consistently, along with all the other engineering about how you handle offline, low Internet, different languages.”
-
Doctors as Entry Point; Bottom-up Adoption (12:55–13:55)
- Compliance and privacy built in from the start. In markets where clinicians/departments have autonomy, Heidi spreads like Slack—users pull Heidi into the system.
-
Enterprise Challenges in the US (13:55)
- “If you build a product for the doctors, then you end up with what Heidi is. If you build it for the CIOs, you end up with just a different thing... more compliant, but less useful.”
(Thomas Kelly, 13:55)
- “If you build a product for the doctors, then you end up with what Heidi is. If you build it for the CIOs, you end up with just a different thing... more compliant, but less useful.”
6. Expanding the AI Experience: Patients and New Use Cases
- Patient Engagement Possibilities (14:49)
- Doctors request features like AI clones to communicate consistent information:
“Can I create a clone of myself in an AI chat?... can you just learn from my sessions how I explain something and then just share it with my patients?”
(Thomas Kelly, 14:49) - “Yes, exactly. Yeah, we can do that.”
(Thomas Kelly, 15:29)
- Doctors request features like AI clones to communicate consistent information:
7. Measuring Success: Capacity and Quality
-
Redefining Healthcare Capacity (15:56)
- “It's more than just time. It's basically how much healthcare is there that we can actually give... If Heidi can...do the job of what the clinician would have done... that's 15 minutes of capacity we've given back.”
(Thomas Kelly, 15:56)
- “It's more than just time. It's basically how much healthcare is there that we can actually give... If Heidi can...do the job of what the clinician would have done... that's 15 minutes of capacity we've given back.”
-
Key Metrics (15:31)
- 51% reduction in documentation time
- 40% improvement in note quality
- 60% decrease in after-hours administrative time
-
Bedside Manner and Human Connection (18:11)
- “Good evidence that bedside manner does matter for outcomes... It’s both. You don’t want to dehumanize the experience.”
(Thomas Kelly, 18:11)
- “Good evidence that bedside manner does matter for outcomes... It’s both. You don’t want to dehumanize the experience.”
8. Location: Pros and Cons of Building Outside Silicon Valley
-
Advantages of Australia (19:18)
- Easier to recruit top talent, unique product-led DNA due to capital constraints, strong work ethic
- “You just can't afford a B2B model generally because you can't raise the quantum of capital that you would need...”
(Thomas Kelly, 19:18)
-
Disadvantages
- Limited domestic market size for global investors.
9. Healthcare’s Future: 10–15 Years Out
-
Vision for a Transformed System (21:10)
- “If I walk into a hospital, it's basically only clinical staff... the administrative requirement to run that hospital has come down a lot... the amount of engagement that patient gets... is massively increased.”
(Thomas Kelly, 21:10)
- “If I walk into a hospital, it's basically only clinical staff... the administrative requirement to run that hospital has come down a lot... the amount of engagement that patient gets... is massively increased.”
-
Prognosis for AI in Care
- AI will handle most primary and routine care, with MDs supervising; AI ensures safety by escalating edge cases and enabling massive scale of patient engagement and monitoring.
Notable Quotes & Memorable Moments
-
On Pivoting During the AI Revolution (10:19):
“At first it was definitely the, oh shit, this is so scary because all of our technical advantage is gone. But then it became exciting...”
(Thomas Kelly, 10:19) -
On Clinical Realities vs. Tech Dreams (12:19):
“You can create any template, you can set up your own voices, high degree of personalization. And Heidi should be writing within 1% of what you would write reliably, consistently.”
(Thomas Kelly, 12:19) -
On Measuring Healthcare Capacity (15:56):
"It's kind of like in physics where you have work. It's like power and time. You want how much care has been given over what period of time."
(Thomas Kelly, 15:56) -
On the Importance of Bedside Manner (18:11):
"There's good evidence that bedside manner does matter for outcomes."
(Thomas Kelly, 18:11) -
On Building Outside Silicon Valley (19:18):
“You get to recruit just amazing people... unique product led talent DNA in Australia...”
(Thomas Kelly, 19:18)
Timestamps for Key Segments
- 00:06: Surgical lessons for building products: "Economy of movement"
- 03:39: Administrative overload prompting the move to build Heidi
- 07:24: What is Heidi Health? AI as care partner
- 10:19: Responding to the arrival of advanced language models; product pivot
- 12:19: Heidi’s focus on deep personalization and quality
- 14:49: Potential for patient-facing AI features ("AI clones" of doctors)
- 15:31: Metrics of Heidi's impact on documentation and administrative time
- 19:18: The advantages and drawbacks of building in Australia
- 21:10: Envisioning the future healthcare landscape with AI as a care amplifier
Rapid Fire: Practical Builder Insights (24:02 onward)
-
Biggest misconception in healthcare products:
"The best product wins."
(Thomas Kelly, 24:20) -
One product snap-fix wish:
"An amazing permissions and billing system... very, very enterprise."
(Thomas Kelly, 24:38) -
Surprising product:
“Grok… it's slowly becoming my preferred consumer AI.”
(Thomas Kelly, 25:12) -
Favorite non-digital product:
“I have this titanium pen that's like milled from a single piece of titanium. That's like a fidget toy slash pen.”
(Thomas Kelly, 25:39) -
Advice to young healthcare builders:
"Not required to be a doctor... Just design partners. You just have to spend time in the environment, find a doctor that’s willing to have you follow them around... be very clear who is paying for it and who's getting the value."
(Thomas Kelly, 25:59)
Episode Highlights: Takeaways
-
Origin Story Matters:
Start from real experience and problems—Dr. Kelly’s journey began with inspiration from a caring, effective GP. -
Throughput, Not Diagnosis, Is the Bottleneck:
Administrative burden, not medical knowledge, is what most limits healthcare capacity. -
Adaptation in the Age of AI Models:
The leap in foundational models pushed Heidi Health to compete as a deeply product- and customer-focused company. -
Personalization Is Table Stakes in Medical AI:
Products must fit seamlessly, matching the individual clinician’s requirements to be truly adopted. -
AI as Human Amplifier, Not Replacement:
The vision is not fewer doctors, but more bandwidth and focus on what only they can do.
This episode offers a nuanced, honest look at what’s truly involved in building and scaling practical AI for healthcare, from the trenches of clinical care to building global teams, with wisdom for any product leader facing regulated, complex environments.
