Nick Karnik | Blog

Notes on AI, search, and building software.

Writing about what I learn building software. Filter by topic below.

No. 001How to Tell If a Model Is Good EnoughLearn why accuracy is a trap, and how precision and recall reveal what really matters when evaluating machine learning models.May 10, 2026
Engineering
7 min read
No. 002The Modern Attack Surface: How Computers Get CompromisedAn overview of how the attack surface has evolved in modern computing, and the ways computers and accounts get compromised today.Jan 26, 2026
Security
17 min read
No. 003A Practical Way to Think About RAGA grounded mental model for Retrieval-Augmented Generation, with two concrete examples and the tradeoffs that actually matter.Dec 12, 2025
AI
7 min read
No. 004How to Build for Speed: What It Actually Takes to Release FastEveryone wants to move fast, but not everyone knows how. Speed isn’t about heroics or skipping QA. It’s about trust in your systems, your telemetry, and your ability to roll back safely. Lessons from years of shipping at Microsoft, Google, Salesforce, Tableau, and startups on what it actually takes to release fast.Oct 23, 2025
Engineering
7 min read
No. 005It's Not the Launch, It's the LandingIn technology we celebrate launches as if they were victories. The moment something goes live there is a demo, a blog post, a slide in a performance review. It feels like success. But a launch is not the finish line. It is only takeoff. The real test is whether the product lands.Oct 2, 2025
Product
3 min read
No. 006How Engineers Can Use AI EffectivelyAI is everywhere in tech conversations. Some people hype it as magic while others dismiss it as overblown. The truth is simpler. AI is a tool. Like any tool in engineering, its value depends on how it is used. Used carelessly, it produces garbage. Used well, it creates leverage.Sep 26, 2025
AI
3 min read