Writing
Long-form notes on the systems underneath the API — Kubernetes, serving engines, agent harnesses, retrieval and the parts of evaluation nobody enjoys. Written to check my own understanding, so they tend to go one level deeper than they need to.
Learn · 69 concepts System design, one idea at a time Open the shelf →- Building Flo: an agentic RAG assistant for Kubeflow, end to end A technical write-up of my Google Summer of Code 2026 work on kubeflow/docs-agent — from indexing issues and code next to the docs, through hybrid retrieval, a self-hosted model and a locked-down edge, to GitOps, an eval on every merge, and a feedback loop that routes bad answers to the team that can fix them.
- Teaching a docs agent to read the repo Notes from turning kubeflow/docs-agent into an agentic RAG system — why documentation-only retrieval fails for infrastructure questions, and what it takes to index issues, code and manifests alongside prose.
- SGLang, or a runtime that remembers An architecture breakdown of the other serving engine — a radix tree over the KV cache, a scheduler that runs a step ahead of the GPU, and a grammar engine that skips tokens it can already predict.
- Context engineering, after the window stopped being the problem Million-token windows did not end context management, they changed what it is. Attention is the scarce resource now, not space — and that changes what compaction, retrieval and sub-agents are for.
- vLLM from the inside An architecture breakdown of the default open-source serving engine — the scheduler that stopped distinguishing prefill from decode, the block pool underneath PagedAttention, and where the CPU overhead went.
- The harness is the product Everyone ships the same models. What separates an agent that works from one that demos well is the loop around the model — context assembly, tool surfaces, error handling and the budget that stops it.
- The two clocks Almost everything hard about serving a large language model comes from one fact — generation runs on two different clocks, and they are not the same speed.
- Learning Resources Collection of Resources to learn different technologies.
- Colour Detection Implementation Details of Colour Detection Project