The Blog.
Insights and practical guidance on AI, modern web infrastructure, and the engineering behind shipping products that last.
An Integration Pipeline That Copes When Tools Fail: Effect, Alchemy and Cloudflare
How we built a live integration pipeline demo in TypeScript: Effect for retries, timeouts, a circuit breaker and typed errors; Cloudflare Workers, Queues and Durable Objects to run it; Alchemy to define the infrastructure in the same program; and Foldkit for the page. The stack, the design, and the code.
Four AI Demos on Cloudflare's Free Tier: How We Built Them
We built four working AI demos (document Q&A, invoice reading, quote drafting, and a shop assistant) on Cloudflare Workers, Workers AI, Durable Objects, Workflows, D1, and AI Search, all inside the free allowances. The architecture, the measured cost of each request, and what we learned.
Cloudflare Security Best Practices for AI Apps
The security practices we follow when we put AI on Cloudflare Workers: server-only session state, a budget in front of every model call, safe handling of uploads and saved results, rate limiting at two levels, and the seven layers between the internet and a model call. Written with a focus on keeping the bill under control.
The Cost-Efficient AI Stack: Ship AI Features Without the Runaway Bill
Most teams overpay for AI by routing every request to a frontier model. This is the architecture we build instead — hybrid cloud+local routing, self-hosted inference, agent orchestration, and cost-per-request observability — and the single principle that ties it together: send each unit of work to the cheapest model that can do it well.