Lenny’s Podcast102 min
Roman Ugarte · Sep 8, 2026
Key learnings- Build a new product from scratch with a small, isolated team rather than retrofitting an existing surface, so each micro-decision…
- Onboard early users by hand and sit in on calls; fixing failures the same day compresses learning dramatically.
- Let users reveal patterns on their own before productizing them, then gently encourage the winning patterns without making them…
Lenny’s Podcast76 min
Heidi Helfand · Jan 18, 2024
Key learnings- Treat reteaming as a people-layer problem to be designed for deliberately, not just a product or delivery concern, since teams…
- Use transparency tools like whiteboards showing proposed team structures, names, missions, and open roles to gather input and…
- Clarify who requests, who gives input, who decides, and who executes each change (the RIDE framework, credited to Pat Wadors) to…
Lenny’s Podcast53 min
Shweta Shriva · Apr 9, 2023
Key learnings- Build autonomous driving to feel natural and predictable, using human driving data while discarding bad driving behavior, so…
- Expect the MVP bar to be far higher when safety is at stake, but still ship early and iterate on real-world deployment.
- Track both commercial/operational metrics (trips, active users, cost) and system behavior metrics (safety versus human…
Lenny’s Podcast142 min
Simon Willison · Apr 2, 2026
Key learnings- Coding agents crossed a reliability threshold around November 2025, so they now usually do what you ask, which changed how…
- Writing code is now cheap, so the bottlenecks move to ideation, prototyping, testing, and process; build several quick prototypes…
- Use test-driven development with agents: have them write tests, run them, and watch them fail first (Red/Green TDD) to keep…
Lenny’s Podcast92 min
Joe Hudson · Aug 8, 2024
Key learnings- Treat the critical inner voice as usually wrong and unhelpful; rather than trying to silence it, respond to it differently…
- Respond to a fearful inner critic with compassion, such as acknowledging it is scared and that you are there with it, instead of…
- Approach emotional change as experimentation and self-discovery, not self-improvement, since failed attempts at self-improvement…
Lenny’s Podcast131 min
Hamel Husain & Shreya Shankar · Sep 25, 2025
Key learnings- Evals are systematic ways to measure and improve an AI application, essentially data analytics on LLM behavior, replacing…
- Start with error analysis by manually reviewing around 100 sampled traces and writing short notes on the first upstream failure…
- Product people with domain expertise should lead open coding; appoint one 'benevolent dictator' whose judgment you trust instead…
Lenny’s Podcast102 min
Chip Huyen · Oct 23, 2025
Key learnings- Focus on what actually improves AI apps, such as talking to users, improving data, writing better prompts, and optimizing…
- Be cautious about committing to newly released technologies that have not been widely tested, since switching away from them…
- Fine-tuning and post-training can shape model behavior a lot, and many teams now focus effort there because base pre-training…
Lenny’s Podcast117 min
Sander Schulhoff · Jun 19, 2025
Key learnings- Prompt engineering remains valuable: Sander argues that bad prompts can drop accuracy near zero on a task while good prompts can…
- Few-shot prompting is the highest-impact basic technique: give the model several examples of what good output looks like, using a…
- Ask the model to decompose a problem by first listing the subproblems it needs to solve, then solving each one before answering…
Lenny’s Podcast67 min
Teresa Torres · Jun 30, 2022
Key learnings- Use the opportunity solution tree: put a measurable outcome at the root, map opportunities as customer needs and pain points…
- Keep opportunities framed in the problem space rather than as disguised solutions; Torres notes most people mistakenly write…
- Collect rich, behavior-based stories in interviews by asking about the last specific instance (e.g., 'tell me about the last…
Lenny’s Podcast93 min
Sri Batchu · Jun 25, 2023
Key learnings- Sequence B2B growth roughly as founder-led sales, first salespeople, low-cost targeted marketing like content and community, then…
- Use early investor and founder relationships as a growth channel: getting influential operators onto the cap table can drive word…
- Make every growth channel driven by technology and data, such as dedicated growth engineers automating sales workflows and…
Lenny’s Podcast94 min
Hari Srinivasan · Jul 16, 2023
Key learnings- Anchor decisions in a single clear North Star, such as connecting people to economic opportunity, so cross-marketplace tradeoffs…
- In complex ecosystems, map second- and third-order effects of every change, since a feature can shift how members perceive and…
- Name one decision-maker per decision (RAPID-style recommender, agreer, decider, input roles) and move to a call after a few email…
Lenny’s Podcast129 min
Fiona Fung · Jun 21, 2026
Key learnings- Manage with AI by running recurring Claude sessions over repos, Slack, and metrics to summarize shipped work and turn it into…
- Automate daily leadership rituals such as scanning feedback channels with scheduled routines that surface themes and draft PRs…
- Give Claude explicit frameworks such as checked-in specs and skills describing what good looks like, so automated code review can…
Lenny’s Podcast65 min
Hamilton Helmer · May 5, 2024
Key learnings- Start thinking about strategy and power even before product-market fit, focusing on the underlying characteristics that might…
- A real power needs both a benefit (cost or price advantage) and a barrier that competitors cannot easily mimic; Helmer calls this…
- Early-stage startups should prioritize counter positioning, then consider network economies, scale economies, and switching costs…
Lenny’s Podcast78 min
Ian Silber · Aug 16, 2026
Key learnings- Designers are still early in adapting to AI; nobody has a settled design process, so starting to experiment today already gives…
- Design work has not sped up as much as engineering because the process still requires messy iteration, trying and discarding…
- Use AI tools across the whole design process, including dropping early ideas into a coding or agent tool to prototype and think…
Lenny’s Podcast70 min
Upasna Gautam · Feb 23, 2023
Key learnings- Build buffer time into plans for newsroom work, since breaking news routinely pulls journalists away from scheduled research and…
- Use recurring touchpoints such as weekly demo days, multi-session working sessions, and office hours to gather continuous…
- Script breaking-news dress rehearsals to stress-test the platform under realistic time pressure before a real event hits.
Lenny’s Podcast117 min
Alexander Embiricos · Jan 12, 2026
Key learnings- Ship a product to a small group first, then learn empirically from usage rather than over-planning, especially in fast-moving AI…
- Onboard users into agentic tools by first working alongside them in the local IDE or terminal, then gradually configure…
- Give AI coding tools your hardest real tasks rather than trivial ones, and build trust by having the agent first understand the…
Lenny’s Podcast58 min
Dylan Field · Jun 30, 2024
Key learnings- Treat intuition as a hypothesis generator: propose hunches, then debate them and test them against data before committing.
- Read customer feedback constantly and dig past stated requests to find the real underlying problem users are trying to solve.
- Ship early to get feedback faster, but hold a firm quality bar, since software can be iteratively improved after launch.
Lenny’s Podcast115 min
Eric Simons · Mar 13, 2025
Key learnings- Eric Simons says deep technology bets can take years to find their market; his team stayed alive by bootstrapping and keeping…
- Simons advises treating spending as a default no until you see real customer pull, and buying software with the goal of cutting…
- Simons notes that when a launch unexpectedly takes off, pricing and infrastructure often break first; Bolt rolled out upgrade…
Lenny’s Podcast131 min
Scott Wu · Sep 8, 2025
Key learnings- Treat an AI agent like a junior engineer: scope it with well-defined tasks rather than open-ended problems, and start with small…
- Run several agents asynchronously in parallel and only step in for the portions needing human judgment, such as scoping…
- Invest up front in setup for the agent: connect repositories, teach it how to run lint and CI, and give it a virtual machine so…
Lenny’s Podcast131 min
Boris Cherny · Feb 19, 2026
Key learnings- Ship with minimal scaffolding and let the model choose tools and order of operations, rather than boxing it into rigid workflows…
- Watch how people hack your product for purposes it wasn't designed for; that latent demand, like data scientists using a terminal…
- Bet on the more general model over time instead of fine-tuning or tiny models, since scaffolding gains of 10-20% often vanish…