Podcast
Founders in Arms
Hosted by Immad Akhund and Rajat Suri · EN
In this weekly series, fellow startup founders Immad Akhund (Mercury) and Rajat Suri (Presto, Lima, and Lyft) explore current events in the world of tech, startup, and policy, offering insights from their distinguished careers and an array of expert guests.
YouTube: youtube.com/@FoundersInArms
Substack: foundersinarms.substack.com
Instagram: instagram.com/foundersinarms
TikTok: tiktok.com/@foundersinarms_
16episodes
Episodes
Newest firstAll episodes
WorkOS's Michael Grinich on Becoming the Enterprise Layer for AI's Biggest Companies
2w ago00:54:40Tap to summarizeMichael Grinich is the co-founder and CEO of WorkOS, the enterprise authentication and identity infrastructure used by Anthropic, OpenAI, Cursor, xAI, and hundreds of fast-growing companies. Before WorkOS, Michael dropped out of MIT, worked at Dropbox, and founded Nihilus — where a painful first experience with enterprise features planted the seed for everything that came next.In this episode, Immad Akhund and Raj Suri sit down with Michael to talk about the SaaS apocalypse thesis, how WorkOS quietly became the enterprise layer for AI's biggest companies, and what it actually takes to build for developers.What you'll learn:Why the SaaS apocalypse narrative gets it completely backwardsHow WorkOS became the default enterprise-ready layer for AI-native companiesThe Stripe parallel: why developer infrastructure compounds the same way payments didWhat a failed first startup taught Michael about idea validationHow keeping a daily idea notebook — volume, not quality — led to WorkOSWhy second-time founders approach conviction and validation completely differentlyThe do-or-die bond between developer tools and their customersHow Michael taught himself enterprise sales after starting as a purely technical founderWhy building for developers is the ultimate boss battle in techWhat AI getting to Renaissance-printing-press level actually means for softwareChapters:(00:00) The SaaS apocalypse thesis — and why Michael thinks it's wrong(01:09) Introducing Michael Grinich — MIT, Dropbox, and the road to WorkOS(05:14) The Stripe origin story and early MIT startup network(07:03) Drew Houston, Dropbox, and what convinced Michael to build(09:05) Founding Nihilus: three maxed credit cards and two days from missing rent(11:00) How to generate startup ideas: volume over quality, the notebook habit(14:05) Finding sticky ideas — the ones you keep coming back to(17:10) Why the energy behind an idea matters as much as the idea itself(20:16) What experience gives you: pattern recognition and a framework for new scenarios(24:05) The moment Michael saw the enterprise auth problem and knew it was real(27:02) How Anthropic, OpenAI, and Cursor ended up as WorkOS customers(31:16) Why WorkOS sits at the security and growth layer for AI companies(35:06) The ultimate boss battle: building developer tools for other developers(39:06) Why developer customers give the best product feedback — and why that's a gift(44:04) The SaaS apocalypse revisited — and what's actually happening to software(47:17) How AI compressed the timeline to enterprise-ready from months to a day(53:03) Tying company value to something durable through technology waves
Transcribe →AI Winners, IPO Hype, and the Future of Engineering Teams With Raj and Immad
3w ago00:25:23Tap to summarizeIn this candid one-on-one episode, Immad and Raj catch up on what's actually happening in tech right now — the AI narratives shifting under everyone's feet, which companies they'd bet on, and how they're thinking about building teams in an AI-native world.What you'll learn:Why Anthropic has taken the AI narrative from OpenAI — and whether that lead will holdImmad's take on whether he'd invest in OpenAI or Anthropic at $800B todayHow Anthropic is growing 3x in revenue in three months — and whether it's even possibleThe new engineering team model: fewer engineers, more autonomy, OKR-driven executionWhy design still matters — and why Mercury embeds designers directly into product teamsHow to time IPO investments: why Raj waits 3-4 months post-listing to buyWhat the SpaceX S-1 signals about the new AI hype cycleWhy Apple is undervalued (or not) — the edge computing argumentHow good Gemini's travel integration actually is (Raj tested it in Tokyo)Why AI real-time translation is still painfully clunky — and what the ideal experience looks likeWhere to find Immad and Raj:[00:00] Data centers in space: skeptical takes [01:02] Anthropic's moment: why the narrative has shifted [02:16] OpenAI vs. Anthropic at $800B: where would you invest? [04:12] Anthropic's 3x revenue growth in 3 months: how is that possible? [06:10] The future of engineering teams in an AI-native world [07:37] Design's role in product: why Mercury still embeds designers everywhere [13:44] SpaceX S-1 and the IPO watch list [14:37] Why post-IPO hype fades and when to actually buy [17:01] Gemini in Tokyo: surprisingly good travel integration [17:43] AI translation fails: what the phoneless experience actually needs [20:06] Apple's AI opportunity and the edge computing bet [22:07] Data centers in space: the only scenario it makes sense [24:19] Xai co-founder exodus and AI researcher retention
Transcribe →The Future of Investing: Data, Signals, and Retail Power
Apr 300:52:28Tap to summarizeGeorge Kailas is the CEO of Prospero AI, a platform helping retail investors make smarter decisions using simplified market signals and data-driven insights.In this episode, George joins Immad and Raj to break down one of the biggest debates in investing today: should you just buy ETFs, or can retail investors actually beat the market?They go deep into how modern markets really work, why retail investors are becoming more powerful than ever, and what most people get wrong about stock picking, AI tools, and “free” trading platforms.What you’ll learn:Why ETFs beat stock picking if you don’t have enough timeHow retail investors now make up a massive share of market movement The biggest mistake investors make: not knowing when to exit Why analyst ratings and price targets often can’t be trusted How platforms like Robinhood actually make money (and what it means for you) The shift from software → data as the real moat in AI Why AI stock-picking tools are dangerous in volatile markets The psychology of investing: why most people need to lose before they learnWhat we cover:00:00 Should You Pick Stocks or Just Buy ETFs? 00:50 Meet George Kailas (Prospero AI) 01:30 Beating the Market with Data Signals 02:15 From Mortgage Models to AI Founder 03:20 Why Data Will Matter More Than Software 04:20 Why People Don’t Trust Analyst Ratings Anymore05:00 Who Is Prospero Actually Built For? 05:45 Value Investing vs Modern Momentum 07:00 The Big Debate: ETFs vs Stock Picking 07:35 The 1-Hour Rule: When You Should NOT Pick Stocks 08:30 Retail Investors Are Driving the Market Now09:30 How to Actually Learn Investing (Without Losing Everything) 10:40 Why Exiting Trades Is the Hardest Skill 11:25 Are Public Markets Really Mispriced?11:55 Why Analyst Price Targets Can’t Be Trusted 13:05 Inside Prospero’s 10 Signals System 14:10 How They Simplify Complex Market Data 15:10 Risk Signals: When to Exit a Trade16:30 How Traders Use Options, Sentiment & Dark Pools 17:30 Are Apps Like Robinhood Good or Bad? 18:10 The Hidden Cost of “Free” Trades 19:30 Why Retail Investors Lose Power Through Brokers20:10 Better Alternatives to Robinhood 21:40 AI, Data, and the Future of Investing 23:00 Why Intent Data Could Change Everything24:40 AI, Layoffs & Wealth Inequality 26:00 The Rise of Crypto Traders & Risk Culture 27:10 Why Some Investors Need to Lose First 29:00 Why AI Tools Are Bad at Risk30:00 Mercury’s Investing Strategy (Simple ETFs) 31:30 Why They Avoid Complexity in Investing Products31:45 Fundraising Journey: From Angels to Crowdfunding 33:00 Lessons from Running a Crowdfund 34:10 When Crowdfunding Actually Works36:00 Mercury’s Acquisition Strategy Explained 38:00 Building an All-in-One Financial Platform41:00 George’s Founder Journey & Early Exit 42:30 From “Sharky” to Self-Aware Leader 43:30 How Meditation Changed His Leadership Style 45:00 Managing Teams: Autonomy, Mastery, Purpose47:00 Long-Term Vision for Prospero AI 49:30 Rapid Fire Begins 49:40 Founder He Admires (Jensen Huang) 50:40 Trends That Won’t Last 51:30 What He Changed His Mind About52:05 Closing Thoughts
Transcribe →Founding Teams: What Works, What Doesn’t — with Andy Chen
Apr 100:38:40Tap to summarizeAndy Chen is the co-founder of Outcast Ventures, an early-stage fund focused on rethinking how founding teams come together. Prior to Outcast, he worked across recruiting and venture capital, including roles at Riviera Partners, Kleiner Perkins, and Coatue, where he was a General Partner. At Outcast, he’s building a talent-first approach to company creation, including a co-founder matching program designed to help founders form stronger teams from the start.What you'll learn:Why choosing a co-founder from your existing network can lead to weaker outcomesThe data behind why strangers can make better co-foundersWhat actually makes a billion-dollar founding teamWhy Andy evaluates the team before the idea when investingThe key ingredients: skill, interest, and timing alignmentWhy solo founders rarely build generational companiesHow AI is enabling a new wave of high-revenue, small-team businessesThe evolution of venture capital — and what might come nextAndy’s unconventional path into venture, including time in government (as shared in the episode)In this episode, we cover:(00:00) Why successful founders struggle to find co-founders (00:28) Introduction to Andy Chen and Outcast Ventures (01:17) Andy’s path into Silicon Valley (03:23) Building Outcast and rethinking founder formation (04:19) Research on co-founder success (and what most people get wrong) (06:25) Why working with your co-founder before can hurt outcomes (07:47) Skill, interest, and timing alignment in founding teams (08:22) Inside Outcast’s co-founder matching model (10:24) Why existing co-founder platforms often fall short (11:23) Talent vs. finance backgrounds in venture capital (13:37) Why the team matters more than the idea (14:47) How venture capital has evolved over time (17:48) Rethinking the “atomic unit” of startups (19:20) AI, enterprise vs. consumer, and new opportunities (24:49) The rise (and limits) of solo founders (27:48) The future of venture in the AI era (30:33) Rapid fire: trends, feedback, and lessons (34:20) Andy’s experience working in government (37:45) Why everyone should try building something
Transcribe →The Long Game: David Rusenko on Building Weebly, Surviving Acquisitions, and Investing in Climate
Mar 2700:52:00Tap to summarizeDavid Rusenko is the founder and CEO of Leap Forward Ventures, a pre-seed and seed climate tech fund investing in energy, deep tech, and the reinvention of industrial processes. Before that, he spent 14 years as co-founder and CEO of Weebly, growing it from a college project to a platform serving tens of millions of small businesses before selling to Square in 2018.What you'll learn:Why Weebly stayed cash flow positive from early 2009 and what that meant for how they built the companyHow David thinks about dilution — and why inefficient spending is where founders actually lose equityThe three headcount breaking points every CEO hits and how your role has to change at each oneWhy small businesses need owned channels and how marketplaces eating their margin is the defining tension in that marketWhat clean tech investing looked like during the Vinod Khosla era vs. how David approaches it nowWhy solar's cost curve looks nothing like oil's over the last 100 years — and what that means for timingHow David thinks about nuclear's role alongside renewablesWhat made the Weebly acquisition to Square work when most acquisitions don'tHow word of mouth drove 80%+ of Weebly's growth and why that's hard to explain to investorsWhy David moved from operating to investing — and what the coach-on-the-sidelines framing means to himIn this episode, we cover:(00:00) Cash flow positivity and dilution(01:08) Introduction to David Rusenko and Leap Forward Ventures(04:11) What Leap Forward Ventures invests in(05:32) Why climate tech goes through investment cycles(07:09) Oil price vs. solar cost curves over 100 years(09:08) Clean tech timing and the dot-com parallel(10:31) David's take on nuclear energy(12:29) Why David moved from operating to investing(13:45) Reflections on the Weebly acquisition(15:13) The small business owned channel problem(17:57) CEO breaking points at 25, 75, and 175 people(20:02) What happens to your jokes at 75 employees(22:55) Designing culture intentionally as you scale(28:18) Keeping politics out of your organization(32:50) Weebly's lowest points and near-death moments(37:27) Bootstrapping vs. VC — David's actual view(40:18) How Weebly grew: mostly word of mouth(43:04) The three phases of an S-curve market(44:13) What made the Square acquisition work(48:30) Rapid fire
Transcribe →The State of Robotics in 2026: Ryan Gariépy on Hype, Reality, and Long-Term Thinking
Mar 1300:55:44Tap to summarizeThis week, we're bringing back one of our most loved episodes on Founders in Arms. Ryan Gariépy is the co-founder and former CTO of Clearpath Robotics and Otto Motors, acquired by Rockwell Automation for $600M+ in 2023. He bootstrapped the company for five years with only $300K in funding, reached profitability in 18 months, and spent 14 years building mobile robotics platforms that became the industry standard for research and industrial automation. What you'll learn: Why robotics is a systems discipline where progress stacks rather than explodes How to bootstrap a hardware company to $10M revenue before raising venture capital Why robotics follows 20-50% sustained growth for decades vs. software's boom-bust cycles The "promise problem" with humanoid robots and why form factor shapes user expectations How manufacturing in Canada (not China) became a strategic advantage for Clearpath Why founders overestimate 2-year progress but underestimate 10-year impact in robotics The real economics of humanoid robots: $20K cost becomes $80K landed price How robotics investment differs from software: less competitive, more defensible Why experience compounds in hardware but expires in software careers Investment criteria for robotics: engineering risk vs. technical risk and go-to-market strategy In this episode, we cover: (00:00) Introduction and live event announcement (03:29) Ryan's background: Clearpath Robotics and Otto Motors (04:06) Building two brands under one company (06:29) The 14-year journey: challenges and non-linear growth (07:11) Bootstrapping robotics when "nobody thought you could make money" (08:17) Reaching profitability in 18 months with research customers (10:28) Building robotics platforms for MIT, universities, and research labs (11:03) Manufacturing in Canada vs. outsourcing to Asia (15:05) Reconnecting after 20 years: the Waterloo entrepreneurship connection (16:17) Working at Kiva Systems (now Amazon Robotics) (18:10) Why robotics is more exciting now than ever in history (19:21) Robotics as systems discipline: no single breakthrough technology (21:22) The overhype cycle and realistic expectations (22:14) Software explodes then crashes; robotics compounds for decades (23:36) Why hardware is harder but more mission-driven (25:27) The talent pool advantage: people irrationally love hardware (27:30) Physical AI and real-world impact beyond software optimization (28:07) Humanoid robots: incredible tech, miscalibrated expectations (32:41) The "promise problem": form factors make promises to users (34:35) Consumer robotics examples: Matic cleaning robot (35:59) Asia leading in restaurant and airport robotics deployment (38:37) Training challenges and precursor technologies needed (39:20) China's role in robotics and humanoid development (41:08) Venture capital structures forcing "ridiculous things" in robotics (42:36) Robotics for entertainment vs. utility as consumer use case (43:52) Imad's robotics investments: Embark, Gecko Robotics, vertical AVs (45:23) Why robotics is less competitive than software (47:21) Operational design domain and technology risk assessment (48:19) The AV journey: Waymo, Zoox, and the importance of experience (49:39) Experience compounds in hardware, expires in software (50:31) Rapid fire: biggest mistake, following gut over charisma (51:47) Founder inspiration: Rodney Brooks (52:20) Uncomfortable feedback at Honda co-op job (53:17) Investment criteria: engineering risk, go-to-market, team understanding
Transcribe →Thumbtack’s Marco Zappacosta on AI, Trust, and the Future of Marketplaces
Mar 600:51:10Tap to summarizeMarco Zapacosta is the co-founder and CEO of Thumbtack, the home services marketplace connecting homeowners with local pros for everything from plumbing to renovation. Started three weeks before Lehman Brothers collapsed in 2008, Thumbtack has grown to over $500M in annual run rate across 17 years of building.What you'll learn:Why Marco believes Thumbtack is still pre-product market fit at $500M in revenueHow AI is shifting Thumbtack from a search engine to a matchmakerWhy word of mouth is still the biggest competitor to every home services marketplace — and how AI finally evens the scoreWhy convenience doesn't win when someone's spending $1,000 and entering your homeMarco's take on practitioners vs. projectors — and why he doesn't trust most AI predictionsWhy AI agents won't disintermediate high-trust marketplacesHow Thumbtack's operating model evolved from Google to Facebook to their own matrix structureWhat's kept Marco going for 17 years — and why he scores zero on neuroticismWhy Marco wants to stay private a little longer before an inevitable IPOWhy AI applied to robotics is overhyped and synthetic biology is massively underratedIn this episode, we cover:(00:00) AI as substitute vs. complement — the flaw in our collective thinking(01:00) Introduction to Marco Zapacosta(02:12) Practitioners vs. projectors on AI(04:14) Real anxiety about AI job loss — engineers at birthday parties(07:21) Why Marco doesn't trust Block's layoff messaging(09:46) How AI is a massive accelerant for Thumbtack(10:02) Why home services is still pre-product market fit at $500M(11:02) Word of mouth is Thumbtack's biggest competitor(12:40) Will AI agents disintermediate marketplaces?(15:17) Why choice still matters in high-trust purchases(17:34) Why humans still want to read reviews themselves(19:15) Thumbtack's origin story — starting 3 weeks before Lehman collapsed(23:16) What's kept Marco going for 17 years(24:42) Entrepreneur parents and raising entrepreneurial kids(30:20) How Marco runs the company — the matrix model explained(35:25) Four co-founders: how responsibilities divided over time(37:02) Is Thumbtack going public?(39:33) The real downsides of being a public company(45:21) Rapid fire: who inspires Marco, what's overhyped, what's underhyped(47:12) The hardest part of leadership is self-awareness, not skills(49:02) Why struggling early builds staying power
Transcribe →What AI Will Actually Do to the Economy with Noah Smith
Feb 2700:42:05Tap to summarizeNoah Smith is a writer and Substack blogger behind Noahpinion, known for his contrarian, data-grounded takes on economics, technology, and geopolitics.What you'll learn:Why the viral Citrini "2028 Global Intelligence Crisis" post moved markets — and whether it should haveThe psychology behind why "AI causes 2008" scared Wall Street more than killer robotsWhy Noah thinks an AI-driven financial crisis is possible but unlikelyHow a productivity boom could paradoxically trigger a mild recession through "sticky prices"Why AI-enabled bioterrorism — not economic disruption — is Noah's biggest fearWhat Block's 4,000-person layoff and Mercury's hiring shifts reveal about AI's real impact on jobsWhy software job losses in 2023-24 may have been driven by uncertainty, not AI capabilityNoah's take on deflation, GDP growth, and where inflation goes from hereWhy intellectual humility has been Noah's biggest edge as a forecasterThe "dinosaur and the meteor" theory — why we're worrying about the economy while a much bigger threat flies overheadIn this episode, we cover:(00:00) The meteor meme — AI's real threat vs. the economy(01:07) Introduction to Noah Smith(02:14) What the Citrini post actually argued(04:30) Why markets missed Covid — and what that tells us about AI(06:17) Why Citrini moved markets: the power of pattern matching to 2008(07:51) Breaking down Citrini's financial crisis domino theory(08:38) Noah's verdict: possible but unlikely(11:04) Block lays off 4,000 — how does AI-driven unemployment play out macro?(17:42) How a productivity boom could cause a recession: sticky prices explained(19:46) Noah's real AI fear: vibe-coded bioweapons(24:55) Has bioterror surpassed China-Taiwan as Noah's top worry?(25:10) The economy today: inflation, deflation, and GDP(28:44) What Mercury's hiring strategy reveals about AI's effect on headcount(31:32) Why software job losses in 2023-24 may have been forward-looking uncertainty(34:38) The threat to blue collar jobs — are truck drivers next?(35:52) Why intellectual humility is Noah's competitive edge(39:26) The meteor meme closing: we created zombie gods for a 2.7% productivity boost
Transcribe →How AI Agents Will Reshape the Web with Parag Agrawal
Feb 2000:55:52Tap to summarizeWe're bringing back one of our most loved episode on Founders in Arms. Parag Agrawal is the co-founder and CEO of Parallel, building infrastructure for the agentic web. Previously CEO of Twitter, Parag now leads a company architecting how AI agents will interact with the open web at orders of magnitude beyond current human scale. Two years after founding in stealth mode, Parallel recently announced a $100M Series B co-led by Kleiner Perkins and Index Ventures.What you'll learn:Why everything built for human web consumption will become irrelevant when agents become the primary usersHow Parallel's APIs enable agents to search, fetch, and monitor the web with unprecedented scale and speedThe evolution from simple tool calls to autonomous sub-agents with real decision-making capabilityWhy the web must transition from "pull" (searching on demand) to "push" (alerting when conditions are met)The new business models needed to compensate content creators in an agent-driven webParag's counterintuitive approach to fundraising: why VC rejections don't sting but customer rejections doThe rational game VCs play that founders misinterpret as genuine enthusiasmWhy Parag believes we're not in an AI bubble—but an overreaction is coming (and it'll be faster than dot-com)How Parallel built quietly for a year before product-market fit arrived with the agent explosionThe operational philosophy of extreme in-person collaboration that shaped Parallel's early cultureIn this episode, we cover:(00:00) Introduction and Parallel's mission (01:02) What Parallel's APIs enable for AI agents (02:43) Practical examples: coding agents, sales automation, research (04:57) The conviction bet on agents before the market existed (10:54) New business models for content in the agentic web (20:22) The $100M Series B fundraise and going public (23:03) Why Parallel built in stealth with carefully chosen early customers (24:55) Current scale and product offerings (30:42) The evolution from tools to sub-agents to push-based web (33:13) Are we in an AI bubble? Parag's nuanced perspective (36:34) The mental models behind fundraising vs customer rejections (38:37) Why VC enthusiasm is rational strategy, not signal (45:37) Biggest career mistake: delaying Twitter's algorithmic timeline (48:28) The compounding cost of six-month delays (50:09) Finding inspiration in "re-founders" like Satya Nadella (51:54) The most rewarding part: watching customers do unexpected things (52:43) In-person culture and the transition to remote-friendly
Transcribe →Building a Services Business in a Tech World with Honey Homes' Vishwas Prabhakara
Feb 1300:53:32Tap to summarizeVishwas Prabhakara is the co-founder and CEO of Honey Homes, a subscription home maintenance service that's reimagining how Americans care for their homes. After spending four years at Yelp running the restaurant business, Vishwas saw firsthand why marketplaces fail for skilled home services—and built a contrarian solution. Now operating across San Francisco, LA, Chicago, Dallas, and Austin with 3,000+ members, Honey Homes creates quality jobs for skilled workers while delivering consistent, reliable home maintenance to homeowners.What you'll learn:Why the marketplace model fundamentally fails for skilled labor and home servicesThe counterintuitive insight behind every successful consumer business (the Airbnb lesson)How Vishwas discovered workers were shocked that "nobody's yelled at me yet" after joining Honey HomesWhy solving both sides of the market—customer experience AND worker quality of life—is essentialThe role of AI in leveling up service workers and automating operations without replacing humansWhy early compromises on hiring and standards compound into major problems laterThe distribution challenge: getting consumers to prioritize chronic home maintenance needsHow altruism, not just incentives, drives consumer referrals and growthWhy companies like Yelp, Peloton, and Lyft deserve more respect for building culturally relevant businessesThe mental model shift required to sell subscription home services vs. one-time fixesIn this episode, we cover:(00:00) Introduction and the respect successful companies deserve (01:12) YC batch memories and feeling "late" to tech trends (03:05) The genesis of Honey Homes and why Immad and Raj invested (04:50) Growing up with a handy dad and discovering the home services gap (06:30) The counterintuitive consumer insight behind Honey Homes (07:03) "Nobody's yelled at me yet"—the worker experience problem (08:11) Why marketplaces don't work for skilled home services (09:48) Hiring only 1% of handyman applicants (14:07) Building trust through consistent quality and W2 employment (19:31) How altruism drives consumer referrals, not just incentives (21:51) Getting AI-pilled at Vinod Khosla's CEO retreat (23:01) Using AI to level up workers and automate operations (27:54) Overcoming the mental model barrier for subscription home services (30:07) The vision compromise lesson: don't settle on quality early (31:44) The critical importance of distribution for consumer businesses (32:26) Why partnerships aren't the answer (yet) for Honey Homes (38:41) Defending Yelp, Peloton, and Lyft against Silicon Valley discourse (42:18) Unit economics challenges in services businesses (47:10) Role models: Jeremy Stoppelman and Ramit Sethi (48:08) Hope that divisiveness is a passing trend (49:35) The daily challenge of building before the world sees it (51:04) Getting feedback about being "unpredictable" and staying in your head (52:33) Bringing people along for the journey in your mind
Transcribe →