In this one-on-one episode, Immad and Raj catch up on what's happening in tech, and what founders should actually be paying attention to right now.

Fresh from DC, Immad shares a surprising disconnect he observed about AI regulation and market sentiment. The conversation moves through whether we're in another bubble, the startup metrics that are being gamed again, and the infrastructure realities that might change everyone's timelines.

They also get into the fundamentals that separate sustainable companies from hype-driven ones: why retention matters more than growth, what to do after raising at a high valuation, and the frameworks they actually use to make spending decisions.

Plus updates on what they're building at Mercury, Tribe, and Lima—and the lessons they're learning along the way.

A candid founder-to-founder conversation about navigating uncertainty and building for the long term.

Dan Wang is a research fellow at Stanford's Hoover Institute and author of "Breakneck: China's Quest to Engineer the Future." After spending six years living in Hong Kong, Beijing, and Shanghai (2017-2023), Dan witnessed China's technology growth, the US-China trade and tech war, Xi Jinping's increasing authoritarianism, and three years of zero-COVID pandemic controls firsthand.

What you'll learn:

1. Dan's framework of "engineering societies vs lawyerly societies" for understanding the US-China competition
2. How China deliberately promoted engineers to power—by 2002, all nine Politburo Standing Committee members had engineering degrees
3. Why the one-child policy and zero-COVID demonstrate the dangers of literal-minded engineering applied to society
4. How America transformed from building the transcontinental railroad and Apollo missions to being unable to fix its subway systems
5. Why lawyers took over American governance in the 1960s and created a self-reinforcing system
6. The stark reality: China builds 500 gigawatts of solar capacity annually vs America's 50, and has 30 nuclear plants under construction vs zero
7. Why China's electricity advantage could determine who wins the AI race—not just better models
8. How American AI leadership is threatened by power constraints and Chinese researchers potentially returning home
9. Why robotics applications of AI matter more than reasoning models for geopolitical competition
10. The dual reality of America: trillion-dollar tech companies exist alongside broken infrastructure that only works for the wealthy
11. Dan's writing process: traveling, eating (twice), reading novels and history, and being deliberately provocative
12. The future of US-China competition in semiconductors, aviation, manufacturing, and whether America's technological lead is sustainable

In this episode, we cover:

(00:00) Introduction and Dan's AI/electricity thesis

(01:15) Dan's journey from San Francisco tech to China analyst

(03:40) Engineering society vs lawyerly society framework

(04:21) Why engineers running governments can be dangerous

(05:46) The one-child policy: designed by a missile scientist

(06:56) China's path from Mao to engineering-focused leadership

(09:51) America's transformation from builder to regulator (1960s shift)

(11:08) Can the pendulum swing back? Housing, transit, and infrastructure failures

(13:12) The self-reinforcing nature of lawyerly societies

(14:12) Yale Law ambition vs Stanford engineering ambition

(16:13) Is there bipartisan consensus on building?

(17:41) Why left and right can't agree on solutions

(19:32) China's engineering design flaws and authoritarian feedback loops

(22:19) US technological advantages: semiconductors, AI, aviation

(23:07) The electricity bottleneck: China's massive power advantage

(24:31) If AI is everything, what should America do?

(26:29) Why Dan doesn't buy the "AI is everything" premise

(27:27) Robotics as the real AI battleground

(29:35) Silicon Valley codes, China builds power plants

(30:37) Anti-AI populism emerging on left and right

(33:41) Dan's meta process: philosophy, eating, traveling, reading, being provocative

(37:20) China's rural infrastructure and redistribution through building

(40:39) Peter Thiel question: acknowledging China's dual reality

(44:54) America's core tension: works great for the rich, broken for everyone else

(46:35) Will China get stuck in the 2010s like Japan in the 1980s?

Iñaki Berenguer is a serial entrepreneur with three successful exits: Pixable (sold to Singtel), Clink (sold to Thinking Phones), and CoverWallet (sold to Aon for $300M). He's now a partner at Flive Ventures, a $100M fund investing at the intersection of AI and healthcare, and president and co-founder of Ipronics, an AI infrastructure company for data centers.

What you'll learn:

1. How Iñaki built CoverWallet from 0 to $100M in premium revenue and 400 employees in just 4 years
2. Why he'd rebuild his 250-person company with only 10 people in the AI era
3. The hidden time cost of scaling teams: 40% of CEO time spent on HR, hiring, and one-on-ones
4. How strategic partnerships with potential acquirers create acquisition optionality
5. Why investment bankers matter: the difference between 3-month and 8-month due diligence timelines
6. The critical mistake of taking common stock vs. preferred in acquisition deals
7. Why "paranoid optimist" is the ideal founder mindset
8. The lifestyle reality check: VC work vs. founder intensity and what actually counts as "high pressure"
9. Reference check strategies that reveal integrity under pressure
10. How luck and timing determine exits more than founders want to admit

In this episode, we cover:

(00:51) Iñaki's journey: three companies, three exits across different industries

(03:21) Why Pixable's "always on" consumer product was harder than enterprise

(09:04) The decision to sell CoverWallet despite investor pressure to keep building

(12:20) Product-market fit doesn't exist in AI: markets change faster than products

(19:43) How Iñaki would rebuild differently: from 250 employees to AI agents

(22:32) The real time cost of hiring: 100 employees = 1,000 interviews

(27:16) M&A lessons: why time kills deals and investment bankers matter

(29:03) Building optionality through strategic partnerships with potential acquirers

(32:37) The fulfillment of building vs. investing: team wins and external validation

(36:23) Why founders struggle to celebrate wins that took years to achieve

(40:25) The "paranoid optimist" mindset: assuming someone is always working harder

(42:14) AI in healthcare: the most underhyped opportunity

(45:20) Comparing entrepreneurial cultures: Silicon Valley vs. New York vs. Europe

(46:16) The biggest mistake: not doing enough reference checks on people

(48:44) What drives founders: proving doubters wrong, not money

Andrew D'Souza is the founder and CEO of Boardy AI, an AI "super connector" that helps founders, investors, and operators make high-value introductions through voice conversations. Previously, Andrew co-founded and scaled Clearco (formerly ClearBank) from a YC Fellowship company to over $100M in revenue and 600 employees across 11 countries before stepping down as CEO to pursue AI innovation.

What you'll learn:

1. How Andrew pivoted Clearco through three different markets before finding product-market fit with e-commerce financing
2. Why customers who "want to be found" create fundamentally easier go-to-market strategies
3. The psychology behind why financial incentives destroy natural networking behaviors and trust
4. How the best companies now generate inbound investor demand instead of running traditional fundraising processes
5. Why Andrew learned to trust founder intuition even when he couldn't articulate it to stakeholders
6. The technical and business model evolution from merchant cash advances to AI-powered networking
7. How Boardy uses voice AI to create more human-like relationship building at scale
8. Strategic insights on building in regulated industries like financial services
9. The transition from scaling a fintech business to creating AI characters with their own objectives
10. Why VCs don't actually remember your previous pitches and how to leverage that reality

In this episode, we cover:

(00:00) Introduction and Andrew's journey from Waterloo to YC

(02:32) Clearco's evolution from Uber driver financing to e-commerce

(04:27) The pivotal board meeting and Series A pivot decision

(05:25) Finding product-market fit with customers who "want to be found"

(09:35) Scaling Clearco to $100M+ revenue and 600 employees

(11:08) The COVID boom and building Clear Angel with GPT-3

(13:31) Andrew's decision to step down as CEO

(15:58) Introduction to Boardy AI and the AI super connector concept

(18:43) Live demonstration of Boardy's voice capabilities

(26:33) Business model and the "economy of Boardy" vision

(29:09) Why financial incentives destroy network effects

(33:36) Fundraising evolution from process-driven to inbound demand

(37:13) The reality of investor relationships and memory

(43:00) Rapid fire: biggest mistakes, inspiration, and founder psychology

(47:29) The creative expression of building AI characters

How do you build a sustainable AI business when investors are throwing money at anything with "AI" in the pitch deck?

We're bringing back one of our best episodes featuring Jesse Zhang (Decagon AI CEO). This is a strategic conversation that centers on the business challenges of building an enterprise AI company that can sustain beyond the current hype cycle.

Jesse reveals how AI moves beyond being a mere chatbot to become a "system of intelligence" that encodes complex business logic, creating moats for companies that implement it effectively. The conversation also explores the realities of implementation, adoption, and enterprise sales strategies that actually work.

Technology founders, customer experience leaders, and investors will find this episode particularly illuminating as it bridges the gap between AI hype and practical implementation. Whether you're evaluating customer service AI, selling enterprise technology, or navigating fundraising, this conversation provides the strategic context and tactical insights needed to make better decisions.

David Gu is the co-founder and CEO of Recall.ai, building conversation recording infrastructure that powers over 1,000 AI companies. Fresh off announcing a $38 million Series B led by Bessemer, David shares the journey from a Winter 2020 Y Combinator call recording tool to becoming the backbone of AI conversation intelligence.

What you'll learn:

How David pivoted from application to infrastructure after spending 80% of engineering time on recording problems

Why the social shift toward recording acceptance created a massive infrastructure opportunity

The systematic approach David used to learn enterprise sales with zero experience

How Recall's desktop recording SDK eliminates the need for bots in meetings

Why Series B fundraising still requires a 100x growth vision even at scale

The framework David uses to validate new products and channels before investing time

How Amanda Gu built 45,000 LinkedIn followers and turned social media into a lead generation engine

Why David records and reviews every sales pitch to improve his closing rate

The mental shift from seeking external validation to embracing continuous failure

How the conversation data revolution will transform every B2B software application

In this episode, we cover:

(00:00) Introduction and David's Y Combinator background

(01:30) Announcing Recall.ai's $38M Series B funding round

(03:18) The pivot from call recording app to infrastructure platform

(09:42) Why recording infrastructure became their nightmare and salvation

(15:36) Learning enterprise sales as a technical founder

(22:13) Amanda's LinkedIn growth and social media lead generation

(28:26) Systematic approach to testing new products and channels

(34:52) Why Series B still requires 100x vision and growth story

(42:17) The social transformation that made recording acceptable

(48:23) Working seven days a week for three years in the early days

(53:05) The framework for embracing failure as a learning tool

Bill Clerico is the founder and former CEO of WePay, which he sold to JPMorgan Chase for $400 million, and is now founding managing partner of Convective Capital, investing in wildfire risk management and physical resilience technologies. Starting WePay during the 2008 financial crisis when VCs said "no one makes money in payments except PayPal," Bill built one of the pioneering fintech companies alongside Stripe and Square.

What you'll learn:

Why VCs avoiding entire sectors often signals the biggest opportunities

The unconventional partnership strategy that led to WePay's $400M JPMorgan acquisition

How to position strategic partnerships as pathways to acquisition rather than just revenue

Why WePay's delayed pivot from consumer to developer APIs cost them market leadership

The specific tactics for getting enterprise buyers excited about acquisition vs. partnerships

How to navigate the early fintech landscape without established banking infrastructure

Why timing strategic decisions matters more than perfecting the original plan

The 12-18 month timeline required for enterprise acquisition conversations

How crisis-driven industries create first-time openings for technology adoption

Bill's contrarian thesis on investing in utilities, insurance, and government sectors

In this episode, we cover:

(00:00) Introduction and Bill's journey from investment banking to entrepreneurship

(08:13) Starting WePay during the 2008 financial crisis in Boston

(12:00) Getting into Y Combinator and the early pivot struggles

(17:37) The acquisition strategy and JPMorgan partnership approach

(24:38) Lessons on founder burnout and sustainable company building

(36:19) Convective Capital's thesis on physical risk management

(42:38) Building an insurance company for high-risk California properties

(46:35) The future of wildfire risk and climate resilience investing

Joris Poort is the CEO of Rescale, a digital engineering platform that provides supercomputing capabilities for engineers and scientists designing rockets, drugs, and computer chips. Starting from Y Combinator in 2012, Rescale has grown over 14 years to serve major aerospace and life sciences companies with over 200 employees and a platform that combines high-performance computing with AI physics capabilities.

What you'll learn:

Why "Default Alive" means cash flow positive, not just having runway or theoretical profitability

How to avoid false trade-offs by asking "why not both?" instead of accepting either/or decisions

Why the biggest startup mistakes are always people, especially bad executive hires

The hidden advantages of grinding through difficult early years versus overnight success

How to maintain frugal company culture while scaling from 4 to 200+ employees

Strategic approaches to long-term R&D investments, including AI physics and Department of Defense contracts

The difference between executives who built systems versus those who just ran them

Why promoting internal talent often works better than external executive hires

How to structure resource allocation decisions to force proper prioritization

The psychology of founder endurance and why some businesses are intentionally harder to build

In this episode, we cover:

(00:00) Introduction to Joris Poort and Rescale

(01:04) Meeting Raj at Paul Graham's place in England

(04:36) What Rescale does: supercomputing platform for engineering

(08:02) How modern AI impacts scientific computing and physics

(12:31) Building for 15 years: the ups and downs of long-term company building

(14:20) The moment of becoming "Default Alive" and what it really means

(16:49) VCs versus founders on spending and growth philosophy

(22:03) Implementing frugal culture and budget discipline at scale

(26:56) The challenge of promoting internal talent to executive roles

(32:28) Interviewing every hire up to 200 people and building relationships

(34:30) Rapid fire: riskiest bets, biggest mistakes, and hard-won lessons

(42:31) AI investments, Department of Defense contracts, and strategic moats

(48:45) Why the hardest path often creates the strongest business

Ian Tien is the co-founder and CEO of Mattermost, an open-source collaboration platform that has evolved from a Slack alternative to essential infrastructure for government and defense organizations worldwide. Starting from a failed HTML5 gaming platform at Y Combinator, Mattermost now serves over 4,000 contributors, employs 120 people, and operates profitably while powering sovereign collaboration for national security agencies.

What you'll learn:

1. How Mattermost pivoted from gaming to defense tech through bottom-up adoption
2. The concept of "sovereign collaboration" and why governments need independent AI systems
3. Ian's framework for hiring executives with "scar tissue" from diverse experiences
4. Why he regrets spending $270,000 on internal swag after taking VC money
5. The unique language and procurement challenges of selling to government customers
6. How open source enables PLG motion in highly secure environments
7. The difference between "butts and seats" contracting vs. commercial off-the-shelf solutions
8. Why most successful founders are in their 30s and 40s, not college dropouts
9. Ian's "daddy daughter days" approach to work-life balance as a CEO
10. The emerging landscape of sovereign AI and national security technology

In this episode, we cover:

(00:00) Introduction and Ian's Tools for Tech Leaders podcast

(04:05) Mattermost's evolution from Slack alternative to sovereign collaboration

(08:09) Bottom-up adoption in national security environments

(14:23) Learning to sell to government customers and navigate defense acronyms

(24:49) Executive hiring philosophy and building long-term relationships

(30:47) The sovereign AI landscape and government investment trends

(38:53) Defense tech opportunities and getting started in the space

(42:57) Palantir's business model and government contracting dynamics

(51:18) Rapid fire: Walt Disney as tech founder inspiration and personal lessons

Arvind Jain is the founder and CEO of Glean, building enterprise AI that connects to all internal company data, and co-founder of Rubrik, which recently went public. Starting as an enterprise search company in 2019, Glean has evolved into what Arvind calls "a more powerful version of ChatGPT inside your company," now approaching 1,000 employees and serving the world's most iconic companies.

What you'll learn:

1. How Glean evolved from search to conversational AI without pivoting by riding AI model capabilities
2. Why making it easy for customers to leave can be a competitive advantage in enterprise sales
3. The challenges of competing against every major software company building similar products
4. How to recruit from FAANG companies when you can't match their compensation
5. The difference between building for established markets vs. creating new product categories
6. Why enterprise buyers are tired of AI overpromising and prefer honest positioning
7. How to stay close to product development while scaling to 1,000 employees
8. The transition from Google engineer to startup co-founder and the skills you have to relearn
9. Why customer success and shared roadmaps are critical for complex AI implementations
10. The decision framework for staying independent vs. acquisition when building at scale

In this episode, we cover:

(00:00) Introduction to Arvind Jain and his journey from Rubrik to Glean

(01:27) Lessons learned building two different types of enterprise companies

(03:09) How Glean works as enterprise search and AI assistant

(06:00) The natural evolution from search to conversational AI

(08:00) Navigating the AI hype cycle and competitive landscape

(11:04) Competition strategy and staying ahead of major software companies

(16:40) Honest positioning vs. overpromising in AI sales

(20:01) Building enterprise reputation and customer relationships

(24:41) Motivation for continuing to build after previous success

(26:56) Staying close to product at scale and avoiding bottlenecks

(31:08) Using Glean internally and rapid iteration cycles

(32:18) The decision to stay independent vs. acquisition

(37:24) Transition from Google to startup founder

(43:45) Rapid fire questions and leadership lessons

(48:31) Uncomfortable feedback about constraining team speed