I write about the engineering behind language models, retrieval systems, and data infrastructure.
Technical deep dives and implementation walkthroughs, from the underlying concepts to working code.
Start here
A cleaner mental model for LLM inference: prompt templating, tokenization, scheduling, prefill, KV cache, decode, and the split between the inference engine and the model.
Read the articleConnect the ideas
Ten articles for software engineers on how RAG systems work and fail. Follow worked examples through parsing, indexing, retrieval, context assembly, grounded answers, evaluation, and security.
Understand how prompts become tokens, how attention and KV cache work, and how serving engines scale inference.
Explore this pathLearn how agents use tools, how MCP connects systems, and how multi-agent workflows coordinate.
Explore this pathBuild local chat applications, RAG pipelines, reranking workflows, and natural-language data tools.
Explore this pathLearn DataFrame operations, SQL tuning, Spark optimization, and structured streaming.
Explore this pathFresh from the notebook
A 64-request live benchmark and 48-call review diagnostic show what memory adds to Jev decisions—and where a confidence gate rejects answers it could have delivered.
Follow a project decision from a planner into Hindsight and back to a fresh coder session. Learn memory banks, retain, recall, reflect, and how to integrate persistent memory into...
Separate predictions, policy, and execution with Jev and LangChain. Follow a refund through middleware, inspect review thresholds, and compare recorded decision latency with...
Carry authorization through retrieval, context, caches, and citations. Define trust boundaries, revocation behavior, and operational controls.
Separate retrieval coverage, context quality, and answer support. Calculate ranking metrics and design reproducible regression gates.
Separate schema validity, citation identity, and factual support; design bounded repair, honest abstention, and streaming behavior around those distinct guarantees.
An ongoing practice
Always learning.
Still building.
