Al Engineering 101 with Chip Huyen (Nvidia, Stanford, Netflix)
October 23, 2025
AI Summary
5 min readChip Huyen, author of AI Engineering and former contributor to NVIDIA's NeMo and Netflix's AI efforts, shares grounded insights from building AI products at scale. Drawing from her work with enterprises, she contrasts common misconceptions—like chasing the latest models or frameworks—with what drives real gains, while explaining key techniques like pre-training, fine-tuning, RLHF, evals, and RAG.
What Improves AI Apps, and What Doesn't
Many fixate on staying current with AI news, picking vector databases, or fine-tuning models, but these yield marginal gains compared to basics like talking to users, preparing better data, writing clearer prompts, and optimizing end-to-end workflows. Huyen notes that optimal vs. suboptimal tech choices often differ little in performance (e.g., 5% at most), and switching unproven tools risks lock-in. The psychology here is distraction: hype pulls focus from user feedback, which reveals outsized improvements. For instance, managers debating agent frameworks should first ask if the gain justifies the effort.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
What you'll learn
- 1 **[00:04:34] Viral table: What people think vs. what actually improves AI apps**
- 2 **[00:07:34] Pre-training vs. post-training and fine-tuning basics**
- 3 **[00:15:21] Supervised fine-tuning, RLHF, and data labeling economics**
- 4 **[00:22:41] Evals: Design, importance, and tradeoffs for apps**
- 5 **[00:32:04] RAG explained: Retrieval-Augmented Generation and data prep wins**
- 6 **[00:39:32] Enterprise GenAI: Use cases, adoption hurdles, productivity measurement**
- 7 **[00:51:33] AI's impact on engineering: System thinking endures**
+ Full timestamped outline available in the app
Show Notes
Chip Huyen is a core developer on Nvidia’s Nemo platform, a former AI researcher at Netflix, and taught machine learning at Stanford. She’s a two-time founder and the author of two widely read books on AI, including AI Engineering, which has been the most-read book on the O’Reilly platform since its launch. Unlike many AI commentators, Chip has built multiple successful AI products and platforms and works directly with enterprises on their AI strategies, giving her unique visibility into what’s actually happening inside companies building AI products.
We discuss:
1. What people think makes AI apps better vs. what actually makes AI apps better
2. What pre-training vs. post-training is, and why fine-tuning should be your last resort
3. How RLHF (reinforcement learning from human feedback) actually works
4. Why data quality matters more than which vector database you choose
5. Why high performers are seeing the most gains from AI coding tools
6. Why most AI problems are actually UX issues
—
Brought to you by:
Dscout—The UX platform to capture insights at every stage: from ideation to production: https://www.dscout.com/
Justworks—The all-in-one HR solution for managing your small business with confidence: https://www.justworks.com
Persona—A global leader in digital identity verification: https://withpersona.com/lenny
—
Where to find Chip Huyen:
• X: https://x.com/chipro
• LinkedIn: https://www.linkedin.com/in/chiphuyen/
• Website: https://huyenchip.com/
• Substack: https://substack.com/@chiphuyen
—
Where to find Lenny:
• Newsletter: https://www.lennysnewsletter.com
• X: https://twitter.com/lennysan
• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/
—
In this episode, we co
More from this podcast
Lenny's Podcast: Product | Career | Growth →