Health Stacks: AI & Tech for Health & Fitness

Mike Hudack (CEO of Sling Money, former executive at Meta, Monzo and Deliveroo) built a personal health data platform almost entirely using Claude Code — connecting his whole genome sequencing, blood tests, Oura Ring, 13 years of Withings data, and workout logs into one AI system. He did it as a side project, with prompts, while running a startup and raising three kids.In this episode, Mike walks through the exact system and how it caught something none of his individual apps could: a specific overtraining period, identified by cross-referencing HRV data with blood tests and workout logs. He also demos a food tracker he built in 10 minutes and explains why he's returning his Whoop.LINKS & RESOURCES- Claude Code: https://claude.ai- Oura Ring: https://ouraring.com- Withings: https://www.withings.com- Strong App: https://www.strong.app- DNAComplete.com: https://www.dnacomplete.com- SNPedia: https://www.snpedia.com- ElevenLabs: https://elevenlabs.io- OpenAI API: https://openai.comTIMESTAMPS0:00 Introduction1:30 Mike's background — Facebook, Deliveroo, Monzo, Sling Money3:02 Building a food tracker in 10 minutes3:59 AI detects overtraining from cross-referenced data5:30 Whole genome sequencing analysis with Claude Code8:23 Why conversational AI beats dashboards9:46 Claude vs ChatGPT as your doctor11:00 Layering data: Oura, Withings (13+ years), Strong app13:30 The FatSecret problem — £1,500/month for UK access15:00 MCP server architecture16:50 Why Mike is returning his Whoop18:00 WhatsApp-based health AI via Open Claude19:24 "Most exciting thing in the last hundred years"20:30 Proactive daily health summaries━━━━━━━━━━━━━━━━━━━━━━ABOUT HEALTHSTACKSAI workflows for training, nutrition, and recovery you can steal today.Hosts: Will Read & James GrantWebsite: https://healthstacks.aiSubscribe for new episodes every Wednesday━━━━━━━━━━━━━━━━━━━━━━#HealthStacks #AIHealth #ClaudeCode #OuraRing #Whoop #GenomicData #HealthData #PersonalAI

James McAulay connected his Garmin, Strava, TrainingPeaks, gym app, and a 65-page blood work report into a single AI coaching system inside Claude Desktop — no code required. He runs four live demos showing exactly how it works. The system critiques recent runs using sleep and HRV data, designs structured workouts based on readiness metrics, spots gaps in gym programming, and personalises post-workout nutrition from specific blood markers. What you'll learn: - How to connect multiple health data sources to one AI agent using MCP servers - A system prompt structure that gives AI enough context to actually coach you - How to turn blood work into personalised, actionable nutrition guidance - The honest limitations — what AI coaching can and cannot replace Resources: www.healthstacks.ai/episodes/ep3-james-mcaulay

James walks through how he built a custom GPT that knows his blood work, health goals, and red lines — and how it recently told him to stop taking glycine.In this episode, you'll see:- How to build a custom GPT that acts as your personal supplement advisor- A prompting technique that compensates for gaps in your own knowledge- Why stacking similar supplements can work against you (and how AI catches it)- How to design 8-12 week experimental review protocols with pass/fail criteria- When to default to your doctor and when AI is the better prep toolTimestamps00:00 Introduction00:30 HealthStacks show intro01:06 Origin story — the scatter-gun approach before AI02:10 Why experimentation loops matter03:21 Key principle: start with goals, not supplements04:02 Screen share: building the Custom GPT06:19 Three health goals explained08:25 System instructions deep-dive08:57 Prompting technique: "ask me clarifying questions"10:14 Getting started advice for beginners14:17 Case study: why James removed glycine17:10 "Top 5 interventions" live demo18:17 Role prompting: specialist vs. generalist21:34 Results walkthrough24:08 Building a full daily stack with dosing schedule27:04 The future of the supplement industry29:13 Failures: when the same AI gives conflicting advice30:34 Final stack review32:50 Wrap-up and takeawaysResourcesChatGPT Custom GPTs: https://chatgpt.comRandox Health: https://www.randox.comGlyNAC: https://examine.com/supplements/glycine/Ashwagandha: https://examine.com/supplements/ashwagandha/Yeast Beta-Glucan: https://examine.com/supplements/beta-glucan/Urolithin A: https://examine.com/supplements/urolithin-a/Website: https://healthstacks.ai#supplements #ai #health #customgpt #biohacking #healthstacks

Stop wasting money on supplements that don't work for you. In this episode, I show you the exact 3-step workflow I use to fact-check health claims, personalize recommendations to my blood work, and find high-quality products at the best price.Will Read demonstrates how he uses Consensus.app to search peer-reviewed research, then ChatGPT with his blood test data to check if supplements are actually right for him — including how AI caught a potential problem with his multivitamin.🔗 RESOURCES FROM THIS EPISODE• Consensus.app: https://consensus.app• Examine.com: https://examine.com• Full show notes: https://healthstacks.ai/episodes/001-bogus-health-claims⏱️ TIMESTAMPS0:00 Introduction1:03 The problem: 15 years of conflicting health advice2:44 Demo: Testing BCAAs in Consensus.app5:17 Red flags that make me reject a supplement6:42 Positive example: Researching creatine8:49 When science disproves what you've heard10:03 Beyond supplements: Training and sauna research12:40 Personalizing with ChatGPT and blood tests14:46 Why iron supplements weren't right for me16:17 AI catching a multivitamin conflict19:11 Shopping for quality supplements22:28 Is all this research worth the time?23:51 The complete workflow🎙️ ABOUT HEALTHSTACKSAI workflows for training, nutrition, and recovery you can steal today. New episodes weekly-ish.🔔 Subscribe: https://youtube.com/@HealthStacksAI🌐 Website: https://healthstacks.ai

Coming Soon: HealthStacks: AI & Tech for Health & Fitness