Notes on building.
Guides on mobile apps, web platforms and AI products. Cost breakdowns, stack comparisons and how we ship.
How to Hire AI Engineers vs ML Engineers for Your Startup in 2026
AI engineers and ML engineers solve very different problems, yet most startups conflate them and waste six figures on the wrong hire. Here is how to tell them apart, what to pay, and where to find the people who actually ship.
AI & strategy · Mar 9, 2026 · 14 min read
From AI Prototype to Production: Shipping LLM Features in 2026
Most AI prototypes never ship. This playbook walks through the exact stages, systems, and org structure required to move an LLM feature from demo to production.
AI & strategy · Mar 9, 2026 · 15 min read
How to Fine-Tune an LLM for Your Domain: A Practical Guide in 2026
Fine-tuning was supposed to be obsolete after RAG. Then LoRA made it cheap and Llama made it accessible. Now it is back as a serious tool for domain accuracy. Here is how to do it without wasting six figures.
AI & strategy · Mar 8, 2026 · 13 min read
AI Content Moderation at Scale: A Technical Guide for Platforms
Every platform with user-generated content discovers content moderation the hard way. AI-powered moderation is now table stakes, but most teams build it wrong. Here is the architecture that actually works.
AI & strategy · Mar 7, 2026 · 13 min read
How to Calculate ROI on AI Features: A Founder's Playbook for 2026
73% of founders cannot articulate the ROI of their AI features. The ones who can raise money faster, ship better roadmaps, and survive board scrutiny. Here is the framework for calculating it honestly.
AI & strategy · Mar 6, 2026 · 12 min read
Self-Hosted LLMs vs API: When to Deploy Llama or Mistral in 2026
Llama and Mistral have closed the gap with GPT-4. Self-hosting can cut LLM costs 70-90% at scale, but only if you can amortize the engineering work. Here is the honest decision framework.
AI & strategy · Mar 6, 2026 · 13 min read
How to Build an AI Red Team: Adversarial Testing for LLM Apps
Every public LLM app gets jailbroken within days of launch. The teams that survive have an internal red team that finds the holes first. Here is how to build one without hiring 12 security engineers.
AI & strategy · Mar 6, 2026 · 12 min read
How to Run LLM Evaluations: A Practical Guide for Founders in 2026
If you ship an LLM feature without evals, you are flying blind. Here is the practical playbook we use to measure, regress, and actually improve production AI systems.
AI & strategy · Mar 5, 2026 · 13 min read
How to Build an AI Phone Receptionist for SMBs in 2026
An opinionated 2026 guide to building an AI phone receptionist for restaurants, clinics, and home services using Vapi, Bland.ai, Retell, Twilio, Deepgram, and ElevenLabs.
AI & strategy · Mar 4, 2026 · 14 min read
How to Build an AI Video Generation Product in 2026
An opinionated, engineering-first playbook for building an AI video generation product in 2026, covering Sora, Veo, Runway, Pika, Luma, Kling, pipelines, GPU costs, and real unit economics.
AI & strategy · Mar 4, 2026 · 14 min read
AI Memory and Context Engineering for Production Agents in 2026
Prompt engineering got you a demo. Context engineering gets you a product. Here is how serious teams actually manage memory for AI agents running in production in 2026.
AI & strategy · Mar 4, 2026 · 14 min read
Build vs Buy AI: A Decision Framework for Startup Founders in 2026
Every founder I talk to in 2026 is stuck on the same question: should we build our own AI or just buy Intercom Fin and move on? Here's the framework I use to decide.
AI & strategy · Mar 4, 2026 · 13 min read