The Cap Table

Adam Seabrook of B Capital discusses AI ROI, valuation shifts, and the rise of value-based care in healthcare investing.

19 min Listen to the original Episode page
Healthcare InvestingValue-Based CareAI in HealthcareVirtual Specialty CarePrivate EquityHealthcare ValuationsGLP-1 Drugs

Summary

Adam Seabrook, partner at B Capital, outlines the firm's focus on healthcare investments, emphasizing a shift from fee-for-service to value-based care models driven by AI efficiencies. He argues that while healthcare lacks the recurring revenue of software, AI can make value-based care viable by improving outcomes and reducing costs. Seabrook highlights virtual specialty care as a key frontier for scaling specialist labor and notes that go-to-market strategies are becoming more critical than technology stacks. He also observes private equity moving earlier into healthcare to acquire assets that can be transformed by AI automation.

Episode sections

  1. 0:08 – 1:01

    Introduction and Background

    Adam Seabrook outlines his background in mechanical and biomedical engineering at Cornell, early startup experience in teleradiology and clinical decision support, and his current role as a partner at B Capital focusing on healthcare.

  2. 1:01 – 2:02

    Daily Operations and Team Scaling

    Seabrook describes his daily routine as centered on learning from new industries and companies. He notes that B Capital scaled its healthcare capabilities by hiring deeply knowledgeable specialists rather than relying solely on individual expertise.

  3. 2:02 – 3:40

    Value-Based Care vs. Fee-for-Service

    Seabrook questions whether value-based care (VBC) can generate venture capital power-law returns due to the financial risk providers must absorb. He argues fee-for-service is historically better but believes AI-driven efficiencies will eventually make VBC viable again.

  4. 3:40 – 4:55

    AI ROI in Healthcare

    He identifies the biggest AI ROI as augmenting clinical teams to improve patient outcomes and reduce friction, rather than just diagnostics. The goal is to free clinicians to provide more empathetic care while lowering costs.

  5. 4:55 – 6:13

    Virtual Specialty Care

    Seabrook argues virtual specialty care is a major frontier because it allows for the amortization of high-cost specialist labor over large, geographically dispersed populations, improving access and outcomes compared to local primary care.

  6. 6:13 – 8:06

    Data Liquidity and Go-to-Market

    He asserts that technology is becoming table stakes, with go-to-market strategy being the key differentiator. Using Picnic Health as an example, he explains how building applications on top of data infrastructure creates defensibility.

  7. 8:06 – 10:33

    Healthcare Valuation Dynamics

    Seabrook contrasts healthcare with software, noting that healthcare lacks recurring revenue models due to curative outcomes and high switching costs. He warns that non-healthcare investors distorted the market in 2020-2021 by applying software multiples.

  8. 10:33 – 18:43

    Private Equity and Cash Flow

    He observes that private equity is moving earlier into healthcare, acquiring businesses that can transform incumbent assets via AI automation, even if those businesses are not yet cash flow positive themselves.

Claims worth verifying

Adam Seabrook co-authored a piece questioning if value-based care can produce power law returns for venture capitalists.

Unidentified speaker ·2:06

How to verify: Search for articles or op-eds authored by Adam Seabrook on B Capital’s website or LinkedIn discussing 'value-based care' and 'power law returns'.

B Capital invested in Synaptic Cure, a company focused on virtual specialty care for neurodegenerative diseases.

Unidentified speaker ·4:56

How to verify: Check B Capital’s portfolio page or press releases for an investment announcement in 'Synaptic Cure'.

B Capital invested in Picnic Health, a company that structures patient medical record data.

Unidentified speaker ·6:16

How to verify: Check B Capital’s portfolio page or press releases for an investment announcement in 'Picnic Health'.

Adam Seabrook was an early employee at Imaging Advantage.

Unidentified speaker ·8:09

How to verify: Check Adam Seabrook’s LinkedIn profile or professional biography for employment history at 'Imaging Advantage'.

Rock Health published a report in 2020/2021 detailing the number of new investors entering the healthcare sector.

Unidentified speaker ·10:12

How to verify: Search Rock Health’s website or annual reports for publications from 2020-2021 regarding investor influx in healthcare.

Revenue cycle management is one of the first areas in healthcare where AI is being used to increase cash flow for existing businesses.

Unidentified speaker ·11:28

How to verify: Review industry reports or news articles from 2025-2026 discussing AI adoption in healthcare revenue cycle management.

GLP-1 drugs are expanding their indications beyond obesity to include chronic disease, dementia, and addiction.

Unidentified speaker ·14:57

How to verify: Search clinical trial databases (e.g., ClinicalTrials.gov) or medical news for recent studies on GLP-1 agonists treating dementia, addiction, or other chronic conditions.

Adam Seabrook lives in Los Angeles.

Unidentified speaker ·17:09

How to verify: Check Adam Seabrook’s LinkedIn profile or public biographies for his location.

San Francisco is considered the tech hub for digital health, despite healthcare being local.

Unidentified speaker ·17:16

How to verify: This is a subjective market opinion stated by the speaker; verification involves noting the speaker's explicit ranking of SF as the 'clear tech hub'.

Wearable data tracking has evolved from limited fitness metrics (Fitbit/Jawbone) to including genomic, gut microbiome, and blood biomarker data.

Unidentified speaker ·13:51

How to verify: Review product specifications and marketing materials from major wearable and diagnostic companies in 2026 to confirm the availability of these data types.

Do your own research

  • Verify B Capital's actual portfolio holdings in value-based care and virtual specialty care companies to assess alignment with Seabrook's stated thesis.
  • Review recent financial filings of major healthcare providers to determine if they are successfully transitioning from fee-for-service to value-based care models.
  • Analyze clinical trial data and FDA approvals for GLP-1 drugs to validate claims regarding their expanding indications into dementia and addiction.
  • Compare valuation multiples of healthcare companies against software companies to assess if the market has corrected the distortions Seabrook mentions from 2020-2021.
  • Investigate the go-to-market strategies of key healthcare AI startups like Picnic Health to evaluate the defensibility claims made by Seabrook.
  • Check private equity deal flow data to confirm the trend of PE firms acquiring earlier-stage healthcare assets for AI transformation.
  • Review independent studies on AI ROI in clinical settings to verify claims about augmenting clinical teams and reducing friction.
Full transcriptexpand
0:00
Unidentified speaker: Hi, I'm Adam Sieberg, and this is the Cap Table.
0:07
Unidentified speaker: Let's get right into it.
0:08
Unidentified speaker: The first question is tell our viewers about yourself from your roots in mechanical and biomedical engineering at Cornell to your mission now as a partner at B Capital.
0:18
Unidentified speaker: So I am Canadian originally, which is relevant as we start to talk about healthcare because I came from a very different healthcare system and was lucky enough to go to Cornell and study engineering there and thought I would develop new prosthetics and ended up getting into the world of startups right out of undergrad and went to work at a teleradiology startup founded by another Cornell grad, worked there for a number of years and worked for a clinical decision support startup as well.
0:47
Unidentified speaker: So two roles at early stage companies and then went to business school at the University of Chicago.
0:54
Unidentified speaker: And when I left, was lucky enough to be hired by B Capital to start working on healthcare.
0:59
Unidentified speaker: All right, next question.
1:01
Unidentified speaker: What does a normal day look like leading healthcare and digital health at B Capital?
1:06
Unidentified speaker: How do you balance the clinical side of evaluating medical breakthroughs with the growth side of scaling a global portfolio?
1:13
Unidentified speaker: Most of our day is spent learning about new things.
1:16
Unidentified speaker: It's new industries, new companies.
1:18
Unidentified speaker: And we have the benefit of having really talented people come in and talk to us about what they're working on.
1:25
Unidentified speaker: And we get the benefit of learning, you know, a career's worth of information in a 30 or 60 minute pitch.
1:32
Unidentified speaker: And so if you like learning, that's a really incredible place to be.
1:35
Unidentified speaker: And that's what we spend a huge portion of our day doing.
1:39
Unidentified speaker: On the other part of the question, the idea of scaling global portfolio with evaluating medical breakthroughs.
1:45
Unidentified speaker: You know, when I started at B Capital, I was the only person focused on healthcare.
1:49
Unidentified speaker: And over time we've built a really phenomenal team.
1:52
Unidentified speaker: And so that's been the component that's allowed us to scale is hiring people who are deeply knowledgeable about healthcare and want to have an impact on the industries that we're working in.
2:02
Unidentified speaker: Next question: Value-based care versus the power law.
2:06
Unidentified speaker: You recently co-authored a piece questioning if value-based care can actually produce power law returns.
2:12
Unidentified speaker: Why is the VC math so much harder for VBC than it is for fee for service models?
2:19
Unidentified speaker: Value-based care implies that the provider is going to take on some amount of financial risk for a patient's care.
2:27
Unidentified speaker: And the piece that I wrote was really focused on the concept of whether or not it was viable for providers to do that and do that at scale in a way that met the returns thresholds that venture capitalists need.
2:40
Unidentified speaker: I think historically it's been hard enough to scale a fee-for-service business and try to do something that's innovative and build a business from scratch without having to take on some new level of risk, like what you have to do in value-based care, where you absorb the financial risk that insurers have historically taken on themselves.
3:03
Unidentified speaker: And so that piece really questioned that.
3:06
Unidentified speaker: And I think my view would be that FIFA service is the better model where you get paid for specifically for the care that you deliver.
3:15
Unidentified speaker: At the end of that piece, uh I started to talk about how AI would be integrated into care delivery and the administrative side of healthcare.
3:22
Unidentified speaker: And so historically, I do think that FIFA service has been a better model than value-based care.
3:29
Unidentified speaker: But as we see efficiencies roll through healthcare, I think you're going to see value-based care come back into fashion and that become a viable model again, which is something that we're excited about.
3:40
Unidentified speaker: Moving right along, in 2026, we moved past the AI hype.
3:45
Unidentified speaker: For healthcare specifically, is AI's biggest ROI in diagnostics and finding the needle, or is it in solving the massive administrative and workflow debt that is hitting hospitals?
3:58
Unidentified speaker: I think that the biggest ROI from AI is going to be getting the most out of the clinical teams that we have in place and delivering better outcomes to patients.
4:08
Unidentified speaker: To me, what that means is using AI to do everything from supporting the diagnosis to identifying or delivering the intervention.
4:20
Unidentified speaker: It means reducing friction and payment, engaging patients.
4:25
Unidentified speaker: And so if we can have AI help with all of those things, I think that frees the clinician up to deliver more empathetic human care.
4:34
Unidentified speaker: And so at the end of the day, you get lower costs, better care for the patient, and a healthcare system that just feels more human, where today I think it feels very cold, your visits are very fast, you move through them and you're in and out of the doctor's office very quickly.
4:49
Unidentified speaker: And so I think if we can deliver on that ROI, that'll be a huge impact for the healthcare system.
4:55
Unidentified speaker: Virtual specialty care.
4:56
Unidentified speaker: With your investment in synaptic cure, you're doubling down on virtual specialty care for neurodegenerative diseases.
5:03
Unidentified speaker: Why do you think specialty care is the next major frontier for virtual health rather than general primary care?
5:10
Unidentified speaker: Primary care is a very local operation, and people say health care is local.
5:16
Unidentified speaker: It makes sense that a lot of our early digital health interventions went towards where a lot of people receive care.
5:21
Unidentified speaker: But as we've expanded into the next wave of virtual care, it's become clear that to deliver the best quality of care to patients, you need specialization.
5:32
Unidentified speaker: And what virtual care allows you to do is get leverage on clinical specialists.
5:38
Unidentified speaker: So you don't have to treat patients locally.
5:41
Unidentified speaker: You can amortize the cost of high-cost clinicians over a large population across geographies.
5:47
Unidentified speaker: You can manage other components of a patient's care through lower cost team members.
5:53
Unidentified speaker: You can coordinate care across a broader group of specialists.
5:58
Unidentified speaker: And so ultimately you end up with, again, lower cost of care for the patient, hopefully better outcomes, and you really increase the access for patients who aren't in large city centers where they might typically have access to these specialists at academic medical centers.
6:13
Unidentified speaker: Moving on, the data liquidity problem.
6:16
Unidentified speaker: Through companies like Picnic Health, you're working on patient-centered data.
6:20
Unidentified speaker: In 2026, is the biggest hurdle to precision medicine the technology itself, or the lack of interoperable data across the healthcare system?
6:28
Unidentified speaker: We talk a lot about how technology is becoming table stakes.
6:34
Unidentified speaker: And one of the major differentiators for companies is actually how they go to market, not their technology stack.
6:41
Unidentified speaker: I think that's true whether you're talking about data interoperability, care delivery, revenue cycle operations, pharmaceutical commercialization.
6:50
Unidentified speaker: So each of those companies is looking for a unique way to go to market.
6:56
Unidentified speaker: In the example of picnic health, before the LL boom came along in 2022, they were building their own AI infrastructure to gather, structure, and abstract data from patient medical records.
7:09
Unidentified speaker: And that's something that's been really hard for healthcare for a long time.
7:14
Unidentified speaker: And so it's a valuable, important piece of technology.
7:17
Unidentified speaker: But very early on, they realized that there would be other people trying to do the same thing and build similar technologies, and that the way that they would stay ahead of the competition is to build applications on top of this really powerful technology.
7:31
Unidentified speaker: And so one example of what they're doing is in clinical trials.
7:36
Unidentified speaker: They can run clinical trials for pharma companies where they are engaged with the patient for much longer.
7:42
Unidentified speaker: So they track patient outcomes for longer.
7:44
Unidentified speaker: They track a deeper level of data than traditional clinical sites, and they can do it at a much lower cost.
7:50
Unidentified speaker: And so they're building defensibility on top of their technology stack by building applications that deliver real value to customers.
7:57
Unidentified speaker: That's how we see companies succeeding, is you need great technology, but ultimately you need a really great business model on top of that as well.
8:06
Unidentified speaker: Moving to the next question: the roll-up strategy.
8:09
Unidentified speaker: You were an early employee at Imaging Advantage, which focused on workflow optimization and rollups.
8:14
Unidentified speaker: What did that experience teach you about the terminal value of healthcare startups that a traditional software VC might miss?
8:21
Unidentified speaker: Working at Imaging Advantage, we went into hospitals and that exposed me to the chaos that exists within healthcare systems in the US in general.
8:32
Unidentified speaker: That is one of the most important learnings is how healthcare actually operates in the US.
8:37
Unidentified speaker: And I think that's something that can be lost on traditional software investors is the difference in the dynamics between selling software and ultimately delivering outcomes to patients.
8:48
Unidentified speaker: So one example is most software startups run on recurring revenue with their customers.
8:56
Unidentified speaker: So you expect those customers to renew year after year, and that brings repeatable revenue, and that's allowed valuations to climb as investors can underwrite more predictable financials for these companies.
9:10
Unidentified speaker: In healthcare, I think we see a lot of the opposite dynamics in some ways.
9:15
Unidentified speaker: So if you are treating a patient and that patient has a curable disease, if you do a great job, you cure the patient and you may never see them again.
9:26
Unidentified speaker: And so that's a good outcome in that case, is you never see your customer again, which is very different than what a software company is hoping for.
9:34
Unidentified speaker: And we see these differences in dynamics in a bunch of different ways in healthcare.
9:39
Unidentified speaker: So Medicare is the one of the largest payers in the country.
9:42
Unidentified speaker: They effectively set their own price for what they're going to pay for services, very different than what happens in software.
9:48
Unidentified speaker: The switching costs are extremely high for a bunch of reasons in healthcare, which means long sales cycles, slower growth.
9:55
Unidentified speaker: All of these things have an effect on cash flow for businesses, which should have an effect on how you value businesses.
10:04
Unidentified speaker: And certainly in 2021 and uh 2020, we saw an influx of non-healthcare investors into the industry.
10:12
Unidentified speaker: And Rock Health put out a great report about the number of new investors who were entering healthcare for the first time.
10:18
Unidentified speaker: And we saw evaluations start to approach traditional software multiples and really distort the market.
10:23
Unidentified speaker: And so I think just some of those dynamics that exist in the operations of healthcare are things that get missed when you don't spend all of your time in the sector.
10:33
Unidentified speaker: Moving right along to the next one, healthcare as a durable asset.
10:37
Unidentified speaker: Private equity is increasingly active in healthcare services.
10:41
Unidentified speaker: At what stage does a digital health company stop being a venture bet and start looking like a cash flow positive asset that a private equity shop would want to acquire?
10:51
Unidentified speaker: I think the question kind of nails it that historically private equity has gotten interested when there's a clear line of sight and a pathway to being profitable and starting to generate cash and no risk of more operational cash needing to go into the business.
11:07
Unidentified speaker: But one of the things that's interesting is we are seeing private equity move earlier in acquiring businesses that can start to transform their more incumbent services heavy assets.
11:18
Unidentified speaker: So looking at AI companies that can really increase the cash flow of existing businesses by automating a bunch of the human labor that goes into them today.
11:28
Unidentified speaker: So we're seeing that in revenue cycle is one of the first areas where that's happened in healthcare.
11:34
Unidentified speaker: We are seeing it in a lot of back office administrative workflows, and I'm certain we're going to start to see it in healthcare delivery more and more.
11:42
Unidentified speaker: And so I would expect a private equity to be buying businesses that are not necessarily cash flow positive, but can really have an effect on the cash flow of the businesses that they may already own or are looking to buy.
11:55
Unidentified speaker: Moving on to the next question: the engineer's conviction.
11:59
Unidentified speaker: You have a background in mechanical engineering.
12:01
Unidentified speaker: How does thinking like an engineer change how you evaluate a clinical founder compared to an investor who comes purely from a finance background?
12:10
Unidentified speaker: Thinking like an engineer is really about breaking down these big problems into smaller and smaller problems, and then trying to get answers and kind of continuing to move down the chain until you've broken the problem down and understand the risk as much as you possibly can.
12:26
Unidentified speaker: Ultimately, what we're trying to do is understand the risk that we're taking in any investment because every early stage company has risk and evaluate whether or not that's a level of risk that we're comfortable taking.
12:39
Unidentified speaker: I think the other component there is not being afraid to ask the dumb question.
12:45
Unidentified speaker: We're looking and to invest in incredibly talented people.
12:50
Unidentified speaker: Most of the founders we've invested in are brilliant and exceptional at what they do.
12:55
Unidentified speaker: And so when you're asking them a question, you're just inherently asking someone who knows more.
13:00
Unidentified speaker: So there's no risk to feeling dumb.
13:02
Unidentified speaker: You are going to be dumb compared to them in that context.
13:06
Unidentified speaker: And so that gives you the freedom to really ask any question that you want and just keep asking the question that everyone might know the answer to and think is obvious.
13:15
Unidentified speaker: Uh, but you can just keep asking until you feel comfortable.
13:18
Unidentified speaker: And so thinking like an engineer uh is really about just continuing to drill down until you have that level of comfort and asking the dumb question if you need to.
13:28
Unidentified speaker: Fill in the blank.
13:29
Unidentified speaker: What's a technology or medical trend that you once thought was hype, but now realize is actually the future of healthcare?
13:38
Unidentified speaker: I think this one is uh wearable data, personal healthcare data that 10 years ago might have been constrained to your Fitbit or jawbone and is now uh much deeper than that.
13:51
Unidentified speaker: So every watch is tracking what you're doing.
13:53
Unidentified speaker: You can get genomic data, you can get gut microbiome data, you can get blood biomarkers from a number of different companies.
14:00
Unidentified speaker: And so I think all of that is interesting because we have better access to the data.
14:04
Unidentified speaker: We are seeing models that are being developed that will help us better understand human biology, and that'll let us make better predictions for uh people's healthcare.
14:15
Unidentified speaker: And so, whereas 10 years ago that felt like people tracking their fitness and maybe more of a fad, it's starting to feel like it'll have a real positive impact on healthcare over the next coming decades.
14:28
Unidentified speaker: Uh, the next section is titled The Lightning Round.
14:32
Unidentified speaker: So uh these are questions I'm supposed to answer quickly, so hopefully I can do that.
14:36
Unidentified speaker: The first one is GLP1s or virtual care, which is the bigger disruptor to the 2026 health economy?
14:44
Unidentified speaker: And I think the answer to that is GLP ones all day.
14:46
Unidentified speaker: We're seeing huge effects in obesity and weight loss, uh, but we're also starting to see where this is gonna have an effect in uh chronic disease and dementia and addiction.
14:57
Unidentified speaker: And so uh I think GLP ones are gonna have a huge impact on uh the healthcare economy, especially as competition brings uh reduced prices.
15:06
Unidentified speaker: That said, I am a huge fan of virtual care.
15:08
Unidentified speaker: I just think GLPs are are so big for healthcare.
15:12
Unidentified speaker: Fee for service, is it actually underrated by VCs right now?
15:17
Unidentified speaker: I wrote that piece saying that I think fee for service is a better model than value-based care typically.
15:22
Unidentified speaker: And so I don't want to say that it is uh overrated, but I think what's actually much more interesting is what's the business model going to be when we see AI integrated into healthcare.
15:33
Unidentified speaker: So fee for service as a model means you are being paid for a unit of your labor.
15:41
Unidentified speaker: And if we start to integrate healthcare AI into healthcare, what we're going to find is that payers will eventually start to pay less if you're using less labor to deliver care.
15:52
Unidentified speaker: That means payers are going to capture all of the value of the efficiency gains.
15:55
Unidentified speaker: I don't think that's right.
15:56
Unidentified speaker: And so I think for providers to capture that gain, they're gonna have to explore new business models like value-based care, bundled payments, uh, those kinds of things that allow them to capture the value of bringing efficiency to healthcare.
16:09
Unidentified speaker: The most important health metric, is it life expectancy or health span?
16:14
Unidentified speaker: Both of these are lagging indicators.
16:17
Unidentified speaker: So if I had to choose one, I would say health span, which is, you know, what's the length of time you have a reasonable quality of life because you're healthy and don't have chronic diseases and other things slowing you down.
16:28
Unidentified speaker: But what I think is going to happen is with the increase in personalized data, you are going to have personalized healthcare metrics that do a better job of tracking you along your personal journey and steering you towards the types of actions that you need to take to avoid specific things that we are predicting based on your healthcare data.
16:51
Unidentified speaker: I think that's much more interesting.
16:53
Unidentified speaker: It probably expands your health span.
16:56
Unidentified speaker: So maybe that's an uh a point in favor of health span, but I think it's going to all be about personalized healthcare metrics relevant to your care training.
17:04
Unidentified speaker: Los Angeles versus San Francisco, which is the true capital of digital health.
17:09
Unidentified speaker: So I live in LA, so I'd love to be able to say LA, but I think San Francisco is clearly the tech hub, and it still is.
17:16
Unidentified speaker: There's a ton of healthcare that's happening here.
17:19
Unidentified speaker: Uh that said, healthcare is local, and we are seeing it in San Francisco, certainly, uh, but in LA and New York and DC and Boston and Chicago and Austin.
17:31
Unidentified speaker: And so uh we're seeing it everywhere, and and we go where the companies are.
17:37
Unidentified speaker: The expensive mistake.
17:39
Unidentified speaker: Uh, what is the biggest error GPs are making in healthcare today?
17:43
Unidentified speaker: Uh, this is a tough one to answer because uh it's all happening so fast and it's playing out in real time, and and different investors have different theses that they're they're investing against.
17:55
Unidentified speaker: What we think is gonna be important for technology companies is finding the right business model and go-to-market strategy.
18:02
Unidentified speaker: And so for new companies, it's gonna be about how do you get your product into customer hands, how do you get paid for what you do, how do you build defensibility?
18:12
Unidentified speaker: And those are gonna be really important questions.
18:14
Unidentified speaker: And so when we look at AI companies, that's what we're looking for on top of technology, which we think is important, starting to become table stakes in a lot of cases.
18:23
Unidentified speaker: Uh, and eventually uh you need to prove that you have a business that will scale.
18:29
Unidentified speaker: And so I don't know uh what the biggest mistake will be, but we think the biggest outcomes will be from companies that have figured out uh what their business model is.
18:39
Unidentified speaker: Thanks to Affinity for having me on the CAT table, and we'll see you on the next one.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3 and interpreted by Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights