Tag
llm
13 pieces tagged “llm”. Browse all tags.
- 11 Oct 2026 Test-time compute, make it work first, and Claude Code with Jira A week covering what test-time compute means, a build-order rule learnt the hard way, and using Claude Code with Jira to keep a backlog clean. Weekly learnings
- 27 Sept 2026 Training an LLM from scratch, and why retrieval beats a better model in coding agents A hands-on LLM-from-scratch tutorial and a Microsoft fastcontext stat suggesting smarter retrieval, not a bigger model, is the biggest lever for coding agent efficiency. Weekly learnings
- 20 Sept 2026 Keycloak, Claude Code's headless OAuth tokens and language model types Exploring Keycloak as an Entra ID alternative, running Claude Code headless with a subscription OAuth token, masked versus autoregressive language models, and why evals matter more than demos when building AI products. Weekly learnings
- 06 Sept 2026 Graphify's token trick, Autocannon load testing and a system design tutorial How Graphify cuts token usage by moving search out of the model, Autocannon for CLI load testing, favourite VS Code themes and a practical system design tutorial. Weekly learnings
- 01 Sept 2026 Chunking strategy for chatbot: stop splitting every 500 characters The default chunking advice is a fixed-size window with overlap. Markdown already tells you where the seams are. Here is the strategy I settled on for my site's retrieval index, why each threshold is what it is, and the failure modes I hit on the way. Blog
- 01 Sept 2026 Teaching my chatbot to read: RAG on Cloudflare's free tier The assistant on this site could name every post I had written and had read none of them. Here is how I replaced prompt stuffing with real retrieval using Cloudflare Vectorize and Workers AI, what it cost (nothing), and the two bugs that ate an evening. Blog
- 30 Aug 2026 LLM gateways, guardrail latency and quality checks in CI Cloudflare Vectorize for the site chatbot, Lighthouse and Lychee as CI quality gates, and AI application lessons on guardrail latency, prompt-injection defenses, gateways, caching risks and chatbot feedback. Weekly learnings
- 22 Aug 2026 Do visitors actually like the chatbot's answers? Opik feedback scores Tracing tells you what your LLM said, not whether it was any good. Adding thumbs up/down to the site chatbot and recording them as Opik feedback scores — with the trace-ID handshake that makes it work. Blog
- 21 Aug 2026 The chatbot on this site: Cloudflare Workers, a $5 budget, and Opik tracing How the 'Ask about my work' chat is built and hosted: a Cloudflare Worker behind the static site, three cost guardrails, and LLM monitoring with Opik. Blog
- 27 Jul 2025 MCP servers and production-grade AI agents Wiring MCP servers into Claude and Cursor, and a blueprint for taking AI agents from dev to production. Weekly learnings
- 20 Jul 2025 FTI architecture, Uber's QueryGPT, Qdrant and a diagram trick Structuring LLM apps with Feature/Training/Inference pipelines, natural-language-to-SQL at Uber scale, a free vector DB for demos, and mermaid → draw.io. Weekly learnings
- 29 Jun 2025 Graph databases, dbt hands-on, AWS Summit — and starting this notes site A packed week: Neo4j vs DGraph with LangChain, a full dbt walkthrough, notes from AWS Summit India, and the first version of this weekly-notes habit. Weekly learnings
- 22 Jun 2025 AWS container services, and tracking LLM benchmarks Comparing ECS, EKS, App Runner and Lambda from a GCP/Azure background, plus a great site for model comparisons. Weekly learnings