Writing · 6 posts
Measurements, and what I changed because of them.
- 20 min
Engineering PlanReel: a 40-page PDF to a narrated video, rendered on 94 Lambdas→
How PlanReel turns a Medicare plan PDF into a narrated video in the agent's own voice, and how moving the render from one small box to 94 Lambdas made it 23 seconds and $0.028.
AI pipelines · Remotion · AWS Lambda · TypeScript
- 12 min
Fine-tuning vs RAG: how LLMs are trained, and what I ship instead→
How an LLM is trained in four stages, when fine-tuning beats RAG, and the third option that solved my own extraction problem: a stock model held to a strict schema.
LLMs · RAG · Fine-tuning · Structured output
- 10 min
Next.js SEO for an AI wedding platform: six fixes, in the order they mattered→
Next.js SEO on an AI wedding-planning platform: headings first, then images, alt text, crawler files and JSON-LD, in the order the work paid off.
Next.js · SEO · Core Web Vitals
- 10 min
Kubernetes vs Single VM: Measured Deploy Time, p99 Latency and the Real Trade→
The same service on one VM and on a three-node cluster, measured with wrk: deploy time, p99 under load, and what you pay for. A trade you can price, not a verdict.
Kubernetes · Benchmarks · Infrastructure
- 9 min
Async/await waterfalls in Node.js: the bug that looks like clean code→
Sequential awaits don't block the Node.js event loop. They add latency you never agreed to pay. How to find async/await waterfalls, and the fix that survives production.
Node.js · Performance · Latency