This week on How I AI: How to create your own AI performance coach & “Farm-to-table software”
Your weekly listens from How I AI, part of the Lenny’s Podcast Network
Hey friends 👋,
Here’s a weekly recap of new podcast episodes across Lenny’s Podcast Network:
Every Monday, host Claire Vo shares a 30- to 45-minute episode with a new guest demoing a practical, impactful way they’ve learned to use AI in their work or life. No pontificating—just specific and actionable advice.
How to create your own AI performance coach: Optimizing nutrition, recovery, and injury management
Brought to you by:
WorkOS—Make your app enterprise-ready today
Google Gemini—Your everyday AI assistant
In this episode, Lucas Werthein, the COO and co-founder of Cactus, breaks down how he built a personal AI health coach that pulls together everything—MRIs, blood tests, wearables, nutrition plans, even journal entries—to help him train smarter, recover faster, and avoid re-injury. After years of surgeries and competing demands on his body, he realized the missing piece wasn’t more data but better synthesis. His AI coach now connects dots across dozens of sources, sets clear performance boundaries, adapts on the fly to real-world changes, and helps him “feel 25 in a 40-year-old body.”
Biggest takeaways:
Data synthesis—not data collection—is the missing link in health optimization. Lucas had access to blood tests, nutritionists, physical therapists, wearable data, and InBody scans, but the challenge was synthesizing this information into actionable insights. His AI coach solved this by integrating all data sources to provide clear, personalized recommendations.
The “last mile” of optimization often comes from unexpected sources. Despite having access to top medical professionals and performance experts, Lucas found that his AI coach unlocked breakthroughs that had eluded him. The AI’s ability to connect dots across siloed data sources provided insights that specialists working in isolation couldn’t deliver.
Effective AI coaches need clear boundaries and anti-prompts. Lucas explicitly instructs his GPT not to push past volume and intensity when metrics show under-recovery, to avoid unproven supplements, and to stick to evidence-based recommendations. These guardrails prevent the AI from giving potentially harmful advice.
AI health coaches will transform the doctor-patient relationship. Within five years, Lucas believes everyone will have access to a personal AI health coach that helps them live healthier between doctor visits and show up more informed. Eventually, a patient’s AI will communicate directly with the doctor’s AI before appointments.
Health-care leaders will eventually sell their knowledge as trained AI models. Lucas predicts that institutions like Mayo Clinic will package decades of patient data into AI models that provide personalized guidance grounded in the best medical science, making expert-level health coaching accessible to everyone.
▶️ Listen now on YouTube | Spotify | Apple Podcasts
“Farm-to-table software”: How I built a Thanksgiving party hub using Lovable for managing invites, dishes, shared recipes, and photos
Brought to you by: WorkOS—Make your app enterprise-ready today
In this episode, Claire walks through how she built a warm, handcrafted “Thanksgiving party hub” using Lovable—and along the way shares a bunch of practical tricks for making AI-generated apps feel intentional instead of generic. She breaks down how she upleveled the typography, visuals, and structure using Google Fonts, Midjourney style references, and a few simple ChatGPT prompts, plus her go-to recipe-cleaning workflow that actually works when you’re cooking with kids.
Biggest takeaways:
“Farm-to-table software” is the new standard for personal projects. Claire’s approach combines multiple AI tools (Lovable for structure, Google Fonts for typography, Midjourney for imagery, ChatGPT for content) to create something that feels artisanally crafted rather than generically generated.
Sometimes the fastest solution is to restart. When ChatGPT 5.1 got stuck processing her recipe request, Claire simply restarted with the “instant” setting rather than waiting—a practical reminder that AI tools sometimes need a reset.
Free Google font combinations are the easiest way to instantly uplevel any design. Instead of accepting default typography, Claire demonstrates how searching “Google font combinations” leads to curated pairings like “Homemade Apple” and “Railway” that immediately transform the look and feel of an application.
▶️ Listen now on YouTube | Spotify | Apple Podcasts
More shows coming soon. . . 👀
If you’re enjoying these episodes, reply and let me know what you’d love to learn more about: AI workflows, hiring, growth, product strategy—anything.
Catch you next week,
Lenny
P.S. Want every new episode delivered the moment it drops? Hit “Follow” on your favorite podcast app.