Why Horizontal Beats Vertical in AI Agents | The Compounding Error Problem Most Founders Miss | The Case For Research-Heavy Teams Win | How to Build AI That Actually Generalizes with Abhishek Das, Co-Founder & Co-CEO of Yutori

The Clay Playbook for Hyper-Targeted Outbound | How to Turn Multiple Signals Into One Story | Why Your List Matters More Than Your Message | The "Magic Wand" Framework for Finding Your Best Customers with Osman Sheikhnureldin, Head GTM Engineering at Clay

Most founders think GTM engineering is just cold outbound done better. Clay's Head of GTM Engineering Osman Sheikhnureldin reveals why that mindset will cost you months of wasted effort. In this episode, Osman walks us through exactly how to identify your ideal customers at the perfect moment, then demonstrates his framework live with two early-stage founders - showing you can't just wing GTM engineering, but when done right, the results compound. Osman Sheikhnureldin is the Head of GTM Engineering at Clay, the company that invented GTM Engineering. He's helped hundreds of startups transform how they use data and technology to remove growth constraints. Joining him are Nilo Rahmani, Co-founder and CEO of Thoras AI (AI-driven cloud reliability and cost optimization), and Panos Papageorgiou, Co-founder of Keragon (HIPAA-compliant automation platform for healthcare). In Today's Episode We Discuss: 01:51 - GTM engineering defined: Solving growth constraints with technology, not headcount 02:57 - Why your target list matters more than your message will ever matter 04:45 - Ditch static ICPs: The jobs-to-be-done framework that actually works 06:53 - The "magic wand" question every founder must answer before building workflows 08:49 - The account scoring workflow no human should ever do manually again 13:46 - How Clay built an ML model to predict contract value from enrichment data 19:18 - Vanta's genius GTM hack using AI screenshots to analyze brand consistency 21:30 - European food startup's signal stack: First US hire + ad spend + new landing pages 23:06 - Early-stage messaging must be hyper-specific—big company tactics won't work for you 26:54 - Founders who lived the pain have an unfair advantage in outbound messaging 28:59 - Counterintuitive truth: AI SDRs have failed—human taste matters more than ever 31:48 - Why hybrid LLM + human skeleton emails crush pure AI-generated copy 34:42 - Voice AI skepticism: Great for extraction, not ready for cold calls 37:18 - Three brutal truths: GTM engineering takes months, hard work, and real creativity 38:20 - "Earn the right to message someone"—the philosophy behind effective outbound 39:24 - Live teardown: Keragon's healthcare GTM using EHR migration as the trigger signal 47:08 - Finding EHR signals through PR announcements, patient portals, and RSS feeds 55:38 - Live teardown: Thoras AI's challenge—spotting cost-cutting triggers that signal growth 01:00:58 - Why "cutting costs" often means a company is scaling, not struggling 01:07:32 - Novel signal: Second product launch as the perfect moment to reach infrastructure teams 01:13:06 - The biggest misconception: GTM engineering goes far beyond cold outbound

Why GTM Engineering is the Future | When to Hire Your First GTM Engineer | How to Treat GTM Like a Product | How Clay Scaled from PLG to Enterprise | Automate the Manual, Never the Important with Yash Tekriwal, Head of Education at Clay

Clay pioneered GTM engineering and went from $1M to $100M in ARR in 2 years. I talked to the person who invented the role of GTM Engineer at Clay. Yash Tekriwal, Clay's first GTM engineer - back when the $3B company was still figuring out what that even meant. What started as one person drowning in too many jobs (RevOps + Sales + BDR + data analyst) has since become a new category that's now reshaping how startups think about go-to-market. You’ll learn: * Why RevOps is "maintenance" but GTM engineering is a growth lever * The skills that define a great GTM engineer today (hint: it involves vibe coding) * What "treating go-to-market like a product" actually looks like in practice * Two org models for GTM engineering teams - and which to start with * "Automate the manual, but don't automate the important" In Today's Episode We Discuss: 01:23 - The origin story of GTM engineering at Clay and why the term is polarizing 05:02 - GTM engineer vs RevOps: maintenance function versus growth lever 07:31 - Treating go-to-market like a product team, not an individual sport 10:42 - Three experiments every GTM team should run on inbound and outbound 15:24 - The essential GTM tech stack: CRM, enrichment, sequencing, and what actually matters 19:08 - Tools founders should consider when getting started—and the automation trap to avoid 22:12 - Zero to $1M: be thrifty on tools and process information manually 25:38 - What to look for in your first GTM engineering hire (hint: it's not technical skills) 28:43 - Signals that you need to hire a GTM engineer for outbound vs inbound motions 31:45 - Scaling past $10M: specialize fast and the hyperscaler dilemma 36:01 - Two org models for GTM engineering: centralized hit team vs embedded engineers 40:22 - The ideal GTM engineer profile: tinkerers, not traditional engineers 43:33 - Why engineers are not the ideal candidates for GTM engineering roles 45:19 - Can salespeople become great GTM engineers? The sales hacker archetype 47:26 - Resources to learn GTM engineering: Clay University, substacks, YouTube channels, and agencies 51:00 - Top three things founders must know about GTM engineering at any stage 52:16 - The most creative GTM engineering builds: satellite imagery, hospital capacity, and custom memes 56:24 - Personal lessons from scaling at Clay: ego death, pivoting, and balancing maintenance with big bets 59:42 - The one thing Yash would change: stop oscillating and let problems become obvious

Why Horizontal Beats Vertical in AI Agents | The Compounding Error Problem Most Founders Miss | The Case For Research-Heavy Teams Win | How to Build AI That Actually Generalizes with Abhishek Das, Co-Founder & Co-CEO of Yutori

The Horizontal vs Vertical AI Debate: Why This Ex-Meta AI Researcher Is Betting Big on Horizontal Web Agents Should you build narrow (vertical) or go broad (horizontal) in AI? This episode unpacks why one PhD researcher abandoned his working vertical product to chase a much riskier horizontal bet - and why VCs leaning heavily into vertical AI might be missing something. Abhishek Das is the co-founder and co-CEO of Yutori, which has raised over $15 million from Radical Ventures, Felicis, and prominent angels including Ali Gil, Sarah Guo, Scott Belsky, and Guillermo Rauch. Previously a research scientist at Meta's FAIR lab, Abhishek holds a PhD from Georgia Tech where he pioneered work on AI agents that can see, talk, and act starting in 2016. In Today's Episode We Discuss: 00:53 - Why how we interact with the web hasn't changed in three decades and what will break that 02:27 - The coming shift from manual browsing to AI assistants performing tasks in the background 05:57 - What "agents" actually meant in ML research before the term became overloaded 06:14 - Why 90% accuracy per step creates catastrophic failure rates over multi-step workflows 08:46 - The behavior pattern humans nail intuitively that machines struggle with: backtracking from errors 10:11 - The DoorDash experiment: building an end-to-end food ordering agent that never shipped 12:58 - Why training on sinle websites leads to memorization instead of generalization 13:03 - The dopamine problem: some tasks users don't want automated 15:08 - Why capability-scoped beats website-scoped: the pivot to read-only horizontal agents 18:05 - Three criteria that drove the horizontal decision: research, user value, and data strategy 24:18 - Scouts API launch: why different channels have different risk appetites for web agents 26:30 - Flying close to the sun: how Yutori competes with hyperscalers on horizontal AI 30:32 - What VCs should actually test for in horizontal AI teams beyond founder horsepower 32:10 - Why three-month roadmaps are the only reasonable planning horizon in AI today 33:05 - The dogfooding ritual: every team member rotates through user feedback weekly 34:50 - Why research and product can't be siloed and how ideas flow both directions 36:03 - The uncomfortable truth: end users don't care about your research breakthroughs 37:32 - The Nintendo Switch 2 problem: aggregating individual feedback into systemic fixes 39:35 - Reframing web agents as "buyer's agents" that filter the internet on your behalf 40:59 - The simulation bet: training agents on cloned websites for high-stakes irreversible actions 43:05 - Why initial team skepticism about Scouts' value proposition was completely wrong 45:01 - How scout reports contextualize results with reasoning and ingest feedback over time 47:52 - The core insight test: where does your instinct lie across research, market, and domain? 49:36 - The hiring trap: why preemptively hiring sales leadership to impress VCs backfires 51:18 - The 12-year-old advice that still guides him: "Be a sponge when entering a new space" 53:05 - Non-negotiables: walking the dog with podcasts and personally reading every user email 54:49 - What founders actually need from VCs: direct and timely feedback, not just capital

Why 90% of B2B Launches Fail | The 10-Week Blueprint That Built 4 Unicorns | Why Different Beats Better in AI Marketing | How to Turn 10 Employees Into 500K Reach | Sumeru Chatterjee, Head of Growth at CoWorker AI

Stop launching into the void and start creating campaigns that actually get attention. Sumeru Chatterjee reveals the exact 10-week playbook he's used to generate millions of impressions at four unicorns, including why most B2B launches fail and how to weaponize your investor network for maximum impact. You'll discover why emotion beats logic in B2B marketing and why treating existing features as "new launches" is the growth hack nobody talks about. Sumeru Chatterjee is Head of Growth at CoWorker AI, the first AI agent for complex work. Previously an early employee at four unicorns and founder of LaunchMob agency, Sumeru orchestrated CoWorker's launch that generated over 1 million impressions in a single day with 75 investors coordinated like a military operation. In Today's Episode We Discuss: 03:45 - Mine 20 customer calls for launch gold using AI 07:29 - Stop hiring humans: Why fear sells better than features 10:23 - The 200 dream customers exercise founders skip 14:50 - Find 50 influencers who already own your audience 18:05 - Three distribution channels max or you'll fail everywhere 20:40 - Everyone claims differentiation but nobody actually commits 29:48 - The $20 book that made Hormozi $60 million 35:36 - Competitive positioning vs contextual positioning changes everything 38:28 - 46 creative concepts: From white papers to doormats 44:16 - Product Hunt is dead for B2B launches 54:05 - Coworker's investor spreadsheet that drove 1M impressions 01:08:44 - Launch every two weeks or become irrelevant 01:11:03 - OpenAI launches Slack integration like it's revolutionary 01:14:37 - Your 10 employees are connected to 500,000 people 01:15:37 - Different is better than better always wins 01:18:04 - Master one funnel before attempting anything else

Why Every AI Company Needs Forward Deployed Engineers | Why You Cannot Build Vertical AI Without Living at the Customer | The 10X Deployment Unlock Nobody Talks About | Pablo Palafox on Building Happy Robot from $0 to Millions in 2 Years

The forward deployed engineer model took Happy Robot from zero to millions in revenue in under 2 years. In this episode, Pablo Palafox reveals exactly how to implement Palantir's FDE motion at a startup - including when to use it, who to hire, and the costly mistakes to avoid. Learn why embedding engineers directly with customers beats traditional sales approaches for complex AI products. Pablo Palafox is the co-founder and CEO of Happy Robot, the AI-native operating system for supply chain and logistics. Prior to Happy Robot, Pablo completed his PhD in computer science and deep learning. Happy Robot has raised over $60M from investors including a16z, YC, and Base10, and serves enterprise customers like DHL. In Today's Episode We Discuss: 02:01 - When the CEO realizes they're actually the company's first FDE 05:13 - Why embedding with customers beats having a sales pitch 09:01 - Stop trying to copy Palantir - build your own FDE playbook 13:25 - The critical difference between FDEs and deployment strategists 17:19 - How Discord servers led to billion-dollar freight broker clients 21:33 - FDEs must become industry insiders, not just tech experts 25:46 - When NOT to use the forward deployed engineer model 30:55 - The FDE hiring secret: Look for "nerds who can sell" 36:35 - Why FDEs need 1-2% equity vs 0.1-0.5% for regular engineers 41:49 - An FDE's day: 80% building, 20% with customers onsite 47:00 - The $30K to $3M expansion playbook for FDE accounts 52:10 - Building FDE "pods" for vertical specialization 56:27 - Why waiting to verticalize FDEs was their biggest mistake 59:19 - The 10X deployment accelerator most startups forget 62:00 - Why FDEs will soon build entire vertical products autonomously 64:58 - When to transition from "things that don't scale" to scale 67:11 - The one place every vertical AI founder must go immediately