Writing
Notes from building the thing, not watching it.
31 pieces on agents, retrieval, model choice and the production engineering that decides whether any of it survives launch.
- 03 Nov 2026Code is no longer the bottleneck: reimagining software engineering for the AI-native SDLCImplementation now runs at agent speed while planning, review, deployment and operations still run at human speed. A walkthrough of The AI-Native SDLC Playbook: from the linear SDLC to a loop driven by committed artifacts.17 min
- 08 Oct 2026Shipping an agentic rewrite behind flags that keep prompts byte-identicalHow to rewrite a working assistant as a tool-calling agent behind flags that leave prompts byte-identical, using shadow mode.7 min
- 08 Oct 2026Disclosed limitations beat confident summariesDesign docs said inputs were sanitized and the code said otherwise. Why drift is normal, and the habit of writing limitations down.5 min
- 08 Oct 2026Never ask the LLM for the headline numberLet the model judge each criterion and let code compute the score. Fixing a match score users could not explain.7 min
- 08 Oct 2026One proposal, one executor, one external writeWhy the model should propose and code should execute: a state machine for external writes, with tamper detection.7 min
- 08 Oct 2026Rulings, not guesses: making \"never assume\" operationalTurning 'never assume' into a pipeline of questions, rulings and records that agents and humans both follow.7 min
- 08 Oct 2026Your status board costs 136K tokens a sessionA coordination process that worked was quietly burning 136K tokens per session on every machine. How we measured it, and the ten rules that cut it.6 min
- 08 Oct 2026Stop lying about what the rule statuses meanA compliance panel that looked real and computed nothing, and how we replaced it with deterministic rules and an LLM fallback.7 min
- 08 Oct 2026Testing LLM systems without a cassette recorder2,804 tests and zero recorded LLM calls: what mocking buys, what SQLite hid, and how to evaluate the model itself.7 min
- 08 Oct 2026The bottleneck is no longer codingA five-point work package went from claimed to delivered in about fifty minutes. What that does to planning, review and the schedule.6 min
- 08 Oct 2026The repo is the only shared stateWith five engineers and five AI assistants, the repository is the only context they all share. What follows from that.7 min
- 08 Oct 2026Three ways LLM JSON fails in production, and the guardrails that stuckA Python dict rendered in the UI, truncated output, and the timeout cliff: three LLM JSON failures and the guardrails that held.7 min
- 08 Oct 2026Thresholds are config, not codeWhy approval limits, tolerances and cut-offs live in a per-tenant config service, and the bugs that taught us to keep two read paths.8 min
- 08 Oct 2026Twelve ways a multi-agent team broke, and the rule each one producedTwelve real incidents from a five-person, five-agent engineering team, and the written rule each one produced.7 min
- 08 Oct 2026Vectors rank. Structured filters decide membership.Cosine scores for right and wrong matches overlap across 0.20 to 0.62. Use vectors to rank and structured filters to decide.7 min
- 09 Nov 2025How AI is Transforming UI/UX Design in 2025Explore how AI tools are empowering designers to work smarter, automate repetitive tasks, and create more personalized user experiences.6 min
- 08 Nov 2025The State of AI Agents and Agentic Software Development in 2025An in-depth look at how AI agents are revolutionizing software development and enterprise operations in 2025.8 min
- 15 Dec 2024Understanding Model Context Protocol (MCP) - The Future of AI IntegrationDiscover Model Context Protocol (MCP), the open standard revolutionizing how AI systems connect to data sources and tools.6 min
- 10 Dec 2024Memory Stores in AI Systems - Building Conversational AI with Long-Term ContextMaster memory stores in AI systems to build intelligent applications with long-term context and personalized experiences.8 min
- 05 Dec 2024Mastering LangChain - Building Production-Ready LLM ApplicationsMaster LangChain to build sophisticated LLM applications with chains, agents, and production-ready architectures.7 min
- 28 Nov 2024Building Multi-Agent Systems with CrewAI - Orchestrating AI TeamsMaster CrewAI to build collaborative multi-agent systems where specialized AI agents work together to solve complex tasks.9 min
- 20 Nov 2024Architecting Single Agent AI Systems - From Simple Bots to Autonomous AssistantsBuild powerful single agent AI systems with advanced reasoning, planning, and tool use capabilities.11 min
- 15 Nov 2024Multi-Agent AI Systems - Orchestrating Collaborative IntelligenceDesign and build sophisticated multi-agent AI systems where autonomous agents collaborate to solve complex problems.13 min
- 28 Jun 2024Implementing Server Actions in Next.jsIn this series, we'll guide you through building a SaaS application using Next.js, Prisma, and Supabase.7 min
- 28 Jun 2024Introduction to Building a SaaS Application with Next.js and SupabaseIn this series, we'll guide you through building a SaaS application using Next.js, Prisma, and Supabase.5 min
- 28 Jun 2024Getting Started with Next.js App Router CapabilitiesIn this series, we'll guide you through building a SaaS application using Next.js, Prisma, and Supabase.5 min
- 09 Jun 2024Understanding the Value of UX in SaaSExplore the essential value of UX and learn how it can transform your SaaS product by enhancing user satisfaction, driving retention, and reducing churn.5 min
- 27 May 2024Emerging Large Language Model ArchitectureA comprehensive guide to the latest advancements in LLMs , including key architectures, techniques, and applications.2 min
- 07 Mar 2024Leveraging RAG for Building a Basic ChatbotHow to leverage RAG for building a basic chatbot, including a step-by-step guide and some code snippets to get you started.4 min
- 10 Nov 2023Architecture antipatternsCommon architecture antipatterns and how to avoid them7 min
- 25 Jan 2023Stability patternsExplanation of common architecture patterns5 min