- 7
- Episodes
- 5
- Ratings
- Weekly
- Cadence
- 2026
- First episode
About Undercovered
Undercovered is about the companies that matter before anyone is paying attention. Each episode, host Marco DeMeireles takes a deeply technical founder to one of New York's best restaurants for a full meal. The conversation goes past the pitch: how the technology actually works, what the hard parts are, what they've gotten wrong, and why they're doing this at all. Marco co-founded Ansa Capital and spends most of his time asking founders questions. Here he does it at the table, for as long as it takes. In partnership with Altamarea Restaurant Group. Produced by Thursday Labs.
- Publisher
- Marco DeMeireles
- Category
- business
- Language
- en
- Explicit
- No
- First episode
- 18 Aug 2026
- Latest episode
- 20 Oct 2026
Latest episodes
7 episodes in the feed.
20 Oct 2026
REVEALED: How AI Finds What Your Doctor Can’t Catch, with Jesse Ehrenfeld @ Aidoc
Tell a healthcare CEO there are thousands of errors in their system, and Dr. Jesse Ehrenfeld says the answer is almost always the same: yeah, we know. Medicine has learned to live with complex, imperfect systems where errors happen. Jesse has spent his career pushing back on that, because he knows what it costs. "I've made errors. I've harmed patients. And those things haunt me." Jesse is the Global Chief Medical Officer of Aidoc, the clinical AI company running in nearly 2,000 hospitals across 20 countries. He's still a practicing anesthesiologist and clinical informaticist. He served as a combat anesthesiologist at the NATO trauma hospital in Kandahar, led Wisconsin's largest health philanthropy, and became the first openly gay president of the American Medical Association before leaving policy to build. Marco DeMeireles sits down with him over an Italian meal at Ai Fiori in Midtown Manhattan to get into medical errors, what it takes for a clinician to trust AI, and why he left the AMA for a startup. In this episode: • Why health systems know about thousands of errors and still can't fix them • Diagnostic errors, and where AI catches what a doctor can miss • Cutting 30 minutes of lag time out of patient care • Would you bet a patient's life on a black box? • The technically impressive tools that nobody uses • Startups vs. the AMA, and what changed when he moved from policy to building • Why his husband talked him into the job he first turned down • The 737 MAX lesson, and why every clinician needs to know where the off switch is • Who should carry the risk when an AI tool gets it wrong • The “human dividend,” and why the best technology stays silent Undercovered is the show where founders building mission-critical companies get the conversation before the world catches up. New episodes weekly on YouTube, Spotify, and Apple. Subscribe so you hear these stories first. Timestamps: 00:00 Intro 01:50 The question he asked the Secretary of Defense in Kandahar 06:01 How AI triage cuts stroke treatment time 12:37 Why he turned the job down the first time 18:04 Startups vs. the AMA 21:46 Would you bet a patient’s life on a black box? 24:19 Technically impressive tools nobody uses 25:29 Who carries the risk when AI misses 27:50 Where’s the off switch? The 737 MAX lesson 30:31 Measuring the human dividend 36:47 Patient safety is AI’s killer app 39:56 Diagnostic errors and 400,000 deaths a year 44:00 What a pill bottle label taught him about impact 49:01 The most undercovered problem in healthcare 52:20 Quick fire round
6 Oct 2026
Why He Said No To 9-Figure Exit Offers, with Brandon Arvanaghi @ Meow
Three weeks after closing a $22 million Series A at a $127 million valuation, Brandon Arvanaghi gave the money back. Not the investors' money, the customers' money, plus every dollar of revenue that had justified the round. Meow brought its revenue down to zero on purpose because he'd heard a rumor he didn't like. Five years later, Meow runs billions in assets and tens of billions in volume, including over $5 billion in stablecoin volume alone, with roughly 40 people, and Brandon has turned down acquisition offers in the hundreds of millions, which he discusses publicly here for the first time. He holds a patent on the first U.S. regulated stablecoin alongside the Winklevoss twins, he made a LinkedIn ad out of a Brex executive's comment about him, and he thinks the best signal in venture is a company with a ridiculous name. Marco DeMeireles sits down with him at Ai Fiori for a conversation about staying alive long enough to get lucky, why hiring is a coin flip and firing is the real skill, and what happens to every fintech margin when agents start shopping on your customers' behalf. In this episode: • Why Brandon took revenue to $0 3 weeks after his Series A • The hiring philosophy behind a 40-person company with billions on platform, and why he says "if you want balance, don't even apply" • Why he thinks hiring is "vibes based" at the end of the day, but Meow's business equation is repeatable • What he's learned from working with Tiger Global and his other investors, and what the best investors do for him • What he learned when a $13 billion competitor publicly admitted they were competing with him • Why agents will compress every commodity margin in fintech, and what stays high-margin • The GEO detail almost nobody knows: agents penalize sites that cost more tokens to read • Why he'd invest in a basket of absurdly named companies without knowing anything else about them • Why he’s turned down recent acquisition offers New episodes of Undercovered every week. Subscribe on YouTube, and follow on Spotify and Apple Podcasts. Episode Timestamps 0:00 The patent next to the Winklevoss twins 3:18 What agents do to customer loyalty 10:09 How Meow actually makes money 12:13 Running at 40 people 13:58 Hiring is a crapshoot. Firing is the skill. 14:52 "If you want balance, don't even apply" 16:22 What's left to compete on in five years 19:16 Five years of refusing to spend 22:45 Tiger, and what great investors do 26:27 Why ridiculous company names win 30:02 Marketing to agents (the token-cost trick) 32:45 The pivot: giving it all back 36:38 The Brex campaign 39:05 The acquisition offers he turned down 43:43 Quickfire
15 Sept 2026
Every Investor Told Him NOT to Build Hardware (But He Did Anyway), with Raj Mehra @ Sage
Every investor who passed on Sage passed for the same reason: "You're awesome, your team is awesome. Why are you building hardware? You could get a great clean seed check if you just build software." Six years and $125 million later, the hardware is the reason Sage has no churn and 180% net revenue retention. Raj Mehra is the co-founder and CEO of Sage, which replaced the nurse call system, a technology essentially unchanged for 40 years, with IoT devices, mobile software for caregivers, and a physical AI layer that went from proof of concept to millions in ARR in six months. Before Sage he failed at a healthcare startup at 24, built the first health system deployment at Palantir, and was an early hire at Cedar. Marco DeMeireles sits down with him at Ai Fiori in New York to walk through every round Sage raised and name exactly what the room got wrong each time. It's the rare founder conversation where the fundraising story is the argument, not the victory lap. In this episode: • Why every investor in the seed round told him to drop the hardware, and how that hardware became the moat • Raj’s formidable failure at Palantir • Why Raj’s last company, Cedar, stalled • Sage's role in the evolution of physical AI • How Raj tested his co-founder’s for fit • The way Sage’s machine learning models were trained • The differentiated work Maveron did at the Series A • How to use debt to creatively fund growth without blowing up the company Undercovered is the show where founders building mission-critical companies get the conversation before the world catches up. New episodes weekly on YouTube, Spotify, and Apple. Subscribe so you hear these stories first. Timestamps: 00:00 Intro 01:59 How he ended up building for senior care 07:10 What Palantir and Cedar each taught him 09:15 His most formidable failure 16:12 The right thesis at the wrong time 19:21 Physical AI and where Sage fits 21:32 The first product, and why nurse call isn't sexy 23:07 Selling with a V0 product and a limited hardware run 25:05 Three co-founders and how they tested each other 28:14 The values they wrote in McCarren Park 33:36 What it costs a 100-bed facility to adopt 34:47 Where the AI actually lives in the product 44:47 Why the hardware made investors nervous 46:01 The investor who toured facilities with her mom 47:43 Using debt instead of more equity 50:49 What investors still miss about this sector
9 Sept 2026
Apathy (Not Disagreement) is the Killer for Your Business, with Alex Konrad @ Upstart
For 12 years, Alex Konrad decided which founders and investors the startup world paid attention to. He ran the Midas List and the Cloud 100 at Forbes, wrote cover stories on Stripe, Zoom, and Snowflake, and met Canva's Melanie Perkins when she was a line item on a rising stars list. Then he left to build Upstarts Media, and found out what it actually costs to be on the other side of the table. Over lunch at Ai Fiori, Marco DeMeireles sits down with a journalist who is now a bootstrapped founder trying to grow a media business in a moment when every piece of advice on his timeline says raise big or die. Alex is candid about the parts nobody writes about: charging readers the price of a sandwich, wanting more complaints in his comments, and the discomfort of hiring people after leaving a place that profited from his own work. The conversation also runs the other direction. Alex turns the questions back on Marco about pricing, diligence, and how investors know if something is medicine or a vitamin. In this episode: - Why Forbes capped Midas List credit at two people, and how firms tried to game it anyway - What it takes to make a reporter open your email, and why success alone doesn't do it - The red flag that makes Alex walk away from a founder before he writes a word - How he found a bootstrapped company with hundreds of millions in revenue that nobody was covering - Why he's paranoid about AI-manufactured narratives showing up as fact - The beat he thinks deserves a full-time reporter before it becomes obvious Subscribe for new episodes every week. Undercovered is on YouTube, Spotify, and Apple Podcasts. Timestamps: 00:00 Intro 01:47 The two founders who come to mind first 03:09 Why he left Forbes to build Upstarts 05:18 Bootstrapping while everyone says raise big 07:12 Apathy is the killer 07:24 The paywall, and charging the price of a sandwich 08:24 How he decides who to cover 09:32 Why he wants more complaints 10:34 Medicine or a vitamin? 12:23 Owning your work after leaving a brand 14:47 How the AI boom compares to the ones he covered before 19:11 Meeting Melanie Perkins on a rising stars list 19:47 How the Midas List data actually worked 21:02 VC egos and who gets credit for a deal 23:46 Why Forbes capped Midas credit at two people 24:13 Rural Pennsylvania and a board-game savant 28:30 The signals that make him lean in 32:43 Non-obvious red flags 34:55 Why he’s paranoid about AI-manufactured narratives 41:32 Quickfire round
1 Sept 2026
"Companies Are Religions, And Religions Need A Bible" | Adam Delehanty, Founder at Ghost
Undercovered is a new show from Ansa Capital Co-Founder Marco DeMeireles, built around one idea: the founders worth talking to are often the ones nobody's heard of yet. Each episode is filmed over a real meal at one of New York's finest restaurants, with a founder building something in defense, cybersecurity, healthcare IT, or another market most people ignore until it breaks. In this episode, Adam Delehanty writes the words other people take credit for. Half his family are writers, lawyers, and journalists. The other half includes a mother who is a social worker and therapist. He describes the job he built out of that inheritance as journalism times therapy: journalism to get the story right, therapy to hold space long enough for a founder to actually deliver it. Ghost is bootstrapped, six years old, and deliberately small. A handful of people, six to twelve clients at a time, a few books a year. He has turned down the obvious growth path more than once, including the one where he takes venture money and builds an army, because he would rather do the work than manage people doing it. The books are the part most people have not heard about. Hinge, Duolingo, DocuSign, Crunchyroll, and one he is not allowed to name. His premise is that companies are religions and religions need a Bible, something on paper a company can orient around instead of a galaxy of Slack threads and Notion pages. To write one he will interview fifty or seventy-five people to find the gap between what a founder believes the culture is and what it actually is. Marco and Adam get into the deltas that keep showing up in those interviews, why he refuses to fill the feed, how he actually uses Claude, the upside-down internet of people who can never post, why he thinks tech reads further right online than it is in private, and the one sentence he wrote that he is contractually unable to claim. Chapters • 00:00 Make something so compelling the world can't ignore it • 02:00 Why founders shouldn't just yap about the product • 06:59 Ghostwriting as journalism times therapy • 07:53 Bootstrapping, and why agencies shouldn't take venture money • 09:14 Companies are religions and religions need a Bible • 10:04 The delta between the culture a founder describes and the real one • 12:46 70% truth, 30% aspirational • 13:15 How the first culture book happened • 14:37 Why the original founder is the best culture tool you have • 18:00 The internet we see and the upside-down internet • 20:04 What happens to tech journalism when every interview is friendly • 22:12 Pricing writing when the floor is falling out • 25:24 AI like a sandwich: human, AI, human • 28:43 What it says about you when you publish slop • 34:45 What VCs get wrong about content • 44:47 Finding the mission in a founder's own story • 50:10 What young writers should do instead of taking the shortcut • 51:22 The hottest take: we can all tell • 52:52 The writer nobody knows, and living an original life • 55:24 Why he thinks tech is more centrist than the feed suggests • 55:58 The famous sentence he can't take credit for
25 Aug 2026
“I left a16z and Raised $41M in 72 Hours” | Zach Cohen, CEO at Ease Health
Undercovered is a show from Ansa co-founder Marco DeMeireles, built around one idea: the founders worth talking to are the ones nobody's heard of yet. Each episode is filmed over a real meal at one of New York's finest restaurants, with a founder building something in a market most people ignore until it breaks. In this episode, Zach Cohen was an investing partner at Andreessen Horowitz, working largely in consumer and consumer health, before resigning to build Ease Health. He flew out to tell his boss he was leaving, and the response was to ask for the deck. a16z led a term sheet inside 24 hours, and the full $41M Series A came together over roughly 72. He has spent the ten months since building three overlapping systems of record at once, CRM, EHR, and revenue cycle, for behavioral health providers who currently run their business across a dozen disconnected vendors. Marco and Zach get into why Zach thinks inpatient psychiatric care, much of it still on paper, is one of the most undercovered problems in healthcare, why he refuses to lead sales pitches with headcount reduction, why he built in New York instead of San Francisco, and what it's like managing four concurrent product roadmaps ten months into a company growing 7 to 8x. Chapters • 00:00 The 24-hour term sheet • 02:10 What it means when the firm you are leaving leads your round • 04:14 From consumer investing to behavioral health • 07:10 Moonlighting in ambulatory clinics for six months before starting • 09:18 Why behavioral health isn't actually crowded • 12:25 What AI-native software can do that legacy EMRs can't • 15:03 Why he may one day give ambient scribing away for free • 16:26 Why build three systems of record at once • 21:51 The "heart surgery" sales pitch • 24:37 Verticals becoming diagonals • 27:56 Why he doesn't lead with labor replacement • 33:51 Choosing New York over San Francisco • 38:16 Hiring the first five, before the round even closed • 43:04 What's in the "fear doc" • 54:15 Growing 7 to 8x, and accelerating into next year • 57:15 The undercovered problem: inpatient psych still on paper • 59:53 His billboard for a billion people
18 Aug 2026
"I Don't Want To Rely On A Distribution Channel, I Want To Be The Distribution Channel” | Taylor Offer, Founder at Atrios
Undercovered is a new show from Ansa Capital Co-Founder Marco DeMeireles, built around one idea: the founders worth talking to are often the ones nobody's heard of yet. Each episode is filmed over a real meal at one of New York's finest restaurants, with a founder building something in defense, cybersecurity, healthcare IT, or another market most people ignore until it breaks. In this episode, Taylor Offer is the first man in four generations of his family not to work in the movie business. His great grandfather, grandfather, uncle, and father all ran film. Then the industry went digital, his dad was a studio projectionist in his 50s with no other skillset, and the family found out what it costs when a technology decides you're no longer necessary. He says the phrase "at the mercy of" three times in this conversation, about three different losses. His father's job. Facebook, which took every dollar of Feat's margin until 5:1 ROAS became 1:1. And the co-founder his board made him fire, a best friend of seven years he spoke to more than his own family. Atrios is his third answer to the same problem: stop renting the distribution channel and become it. This time the channel is other people's trust. Marco and Taylor get into the Google Earth search that found the Vine house, the Logan Paul sock deal that pulled in 20,000 customers, what a private equity sale actually looks like from the inside, cutting a third of the team, and what he thinks the most under covered problem in tech is. Chapters 00:00 "You die when you stop believing you could do it" 02:21 Four generations in the movie business 04:41 "I will never be at the mercy of a job" 05:13 The ten-dollar coffee, and what humans still get paid for 06:29 Building Feat to $50 million 07:29 A hundred cold calls a day, and an app called Vine 08:04 Finding the influencer house on Google Earth 08:46 The Logan Paul sock deal 10:17 Scaling on Facebook 11:12 When five-to-one ROAS became one-to-one 11:27 Deciding to sell 12:48 The accidental discovery of the tastemaker 13:08 What founders don't know about a PE sale 14:10 Cutting a third of the team 17:50 Why Atrios exists 18:13 Is the SDR role dead 20:12 A CRM for tastemakers 23:41 Whether venture is the right partner 26:51 Building without a co-founder 28:48 The board made him fire his best friend 33:44 "More wars are lost by loss of hope than loss of blood" 35:17 The under covered problem: nobody knows how to sell 37:54 The people earning $10-20k a month 54:10 Where to find Taylor
Reach and audience
Public platform figures. Ratings count people who left a rating, not total listeners.
- Apple Podcasts (US)
- 5.0 / 5
- 3 ratings
- Spotify
- 5.0 / 5
- 2 ratings
Contact Undercovered
- Guest appearances
- Books guests
Based on episode analysis; this does not confirm that the show is currently accepting guests.
Host of Undercovered?
Claim your podcast to manage its listing and keep your show details accurate.
Pod Engine is an independent podcast discovery and analytics service and is not affiliated with or endorsed by this podcast. Artwork and show content belong to their owners. Full legal notice.
Explore this show
Podcast research with Pod Engine