05Technical field notes
Practical guides on retrieval-augmented generation, AI agents, LLM evaluation, prompt engineering, embeddings and deploying LLM applications. Written to be read with an editor open.
Showing 9 of 9 articles
FundamentalsGuide8 min read
When an LLM's output feeds other code, treat it like any untrusted input: define a schema, validate strictly, retry with specific feedback, and fail loudly when the model can't comply.
By Alex Morgan
EvaluationGuide11 min read
A one-line prompt edit or a model upgrade can quietly break behavior that used to work. This guide shows how to build a golden set, combine deterministic and model-graded checks, and run them in CI so changes ship with evidence.
By Jordan Lee
RAGExplainer9 min read
How you split documents decides what your retriever can find. This explainer compares fixed-size, recursive, and structure-aware chunking, and shows how to evaluate the choice.
By Priya Raman
DeploymentField note9 min read
When an LLM feature misbehaves, the first question is always what exactly the model saw. These field notes cover what to record per call, how to structure spans across a pipeline, and how to keep sensitive data out of your telemetry.
By Sam Okafor
AgentsGuide10 min read
An agent is a loop that lets a model choose actions. This guide covers how to bound that loop, constrain tools, validate arguments, and design an explicit path for when things go wrong.
By Maya Chen
DeploymentGuide10 min read
Latency and cost in LLM systems are driven by a short list of levers: tokens in, tokens out, which model runs, and how often you call it. This guide walks through each lever and the trade-offs that come with pulling it.
By Sam Okafor
EvaluationChecklist8 min read
A repeatable way to decide whether your LLM feature is getting better or worse: define quality, build a representative eval set, layer cheap and expensive checks, and read results honestly.
By Jordan Lee
FundamentalsExplainer9 min read
Models do not read characters or words; they read tokens, and every request has a fixed token budget. This guide shows how to measure that budget, divide it deliberately, and degrade gracefully when content does not fit.
By Alex Morgan
RAGGuide9 min read
Retrieval and fine-tuning solve different problems. This guide breaks down what each one actually changes, where each tends to fail, and how to decide with evidence instead of instinct.
By Priya Raman
Build the foundations, projects, and habits needed to design more dependable AI applications.