Tom F - Writing https://tomtomtech.net/writing Notes on retrieval, evals, agent infrastructure, and shipping systems that keep working. en-us A refusal-eval rubric for grounded document QA https://tomtomtech.net/writing/refusal-eval-harness https://tomtomtech.net/writing/refusal-eval-harness Four buckets, one table, and why "correct" is the least interesting score in the set Sun, 05 Apr 2026 00:00:00 GMT ai rag evals refusal System-design notes from shipping hosted LLM workloads https://tomtomtech.net/writing/shipping-hosted-llms https://tomtomtech.net/writing/shipping-hosted-llms Prompt caching, tool-use loops, structured output, and the failure modes that show up in regulated deployments Sun, 22 Mar 2026 00:00:00 GMT ai system-design tool-use Notes on designing a protocol-native agent eval substrate https://tomtomtech.net/writing/agent-eval-harness https://tomtomtech.net/writing/agent-eval-harness Why the wire protocol is the right place to draw the eval boundary Sat, 28 Feb 2026 00:00:00 GMT ai agents a2a evals What it actually takes to run unattended for a year https://tomtomtech.net/writing/unattended-for-a-year https://tomtomtech.net/writing/unattended-for-a-year The single design constraint that reshapes every other decision Fri, 20 Feb 2026 00:00:00 GMT systems reliability ops Rust, Go, and the distribution-vs-guarantees matrix https://tomtomtech.net/writing/rust-go-distribution-matrix https://tomtomtech.net/writing/rust-go-distribution-matrix When Rust is the right choice, when Go is, and why the honest answer is "it depends on which problem you're solving" Sun, 18 Jan 2026 00:00:00 GMT rust go systems language-choice