15 articles · from production, not theory
Why AI Agents Fail in Production, and the Architecture That Fixes It
Why AI agents fail in production and the architecture that prevents it: the harness, deterministic guardrails, approval gates, sandboxes, engineered context, evaluation and orchestration, with code and lessons from building production agent systems.
2026-10-06 · 12 min read
Build vs Buy an AI Agent: An Honest Decision Framework
When an off-the-shelf AI tool beats a custom agent, when it doesn't, and the four questions that decide it, from someone who profits from 'build' but will tell you to buy.
2026-07-26 · 4 min read
LangGraph vs CrewAI (2026): Which One to Ship to Production?
LangGraph vs CrewAI, compared on state, durability, human-in-the-loop and multi-agent design, with a clear verdict on which to use for production agents.
2026-06-09 · 4 min read
Designing Agent Architecture with Memory: A Framework from Anthropic's Patterns and LangGraph's Primitives
Anthropic defines how to structure agent workflows. LangGraph provides the memory primitives. Nobody connects the two. This post does, a practical design framework for matching each Anthropic agent pattern to the right LangGraph memory architecture.
2026-05-18 · 15 min read
Adding Memory to Production AI Agents: Mem0, Zep, and LangMem Compared
Your LangGraph agents forget everything the moment a session ends. Here's how to fix it, a production guide to agent memory architectures, Mem0 vs Zep vs LangMem benchmarks, token cost reduction, and GDPR compliance.
2026-05-17 · 12 min read
Production Agent Memory: Compaction, Decay, and the Observation Engine
Most agent memory systems store facts and retrieve them. Production memory does more: it learns behavioral patterns, promotes them to rules, compacts history without losing signal, and scores retrieval with tier-aware decay. Here's the full architecture.
2026-05-08 · 18 min read
How Much Does It Cost to Build an AI Agent? A Developer's Breakdown
Real cost breakdown of building AI agents, LLM API pricing, infrastructure, vector stores, and development time from 4 production agents.
2026-04-03 · 9 min read
GDPR-Compliant AI: Building Guardrails for EU AI Act Readiness
How to build AI systems that comply with GDPR and the EU AI Act, PII filtering, human-in-the-loop, audit logging, data minimization, and LangChain guardrails middleware.
2026-04-02 · 8 min read
How to Build a SaaS with AI Agents: Architecture for Founders
A practical architecture guide for founders building AI-powered SaaS, from agent design and multi-tenancy to billing, deployment, and the build-vs-buy decisions that matter.
2026-03-22 · 7 min read
Prompt Engineering for Production: Beyond ChatGPT Tricks
Production prompt patterns that go beyond tips and tricks, system prompt architecture, structured output, few-shot calibration, context engineering, and prompt versioning with LangSmith.
2026-03-10 · 9 min read
MCP (Model Context Protocol): Connecting AI Agents to Real Tools
Practical guide to MCP, connecting AI agents to tools, databases, and APIs. Covers primitives, FastMCP servers, tool annotations, elicitation, and security.
2026-02-14 · 9 min read
Tokenizing Real Estate on Algorand: A MiCA-Compliant Architecture
Building BitBricks, fractional real-estate tokenization on Algorand with MiCA compliance, ASA RBAC, ARC-20 Smart ASA, and KYC-gated whitelisting.
2025-12-15 · 7 min read
LangSmith for AI Agents: Tracing, Evals & Debugging in Production
A practical LangSmith setup for production agents: tracing every tool call, tracking token cost, offline and online evals, and alerts that catch regressions before users do.
2025-11-22 · 5 min read
LangGraph AI Agents in Production: 5 Patterns That Hold Up
Durable execution with checkpointers, human-in-the-loop interrupts, @task side effects, and multi-agent orchestration, 5 LangGraph patterns from real production agents.
2025-09-18 · 6 min read
Building RAG Pipelines That Actually Work: GraphRAG vs Vector RAG in Production
Vector RAG vs GraphRAG for production, chunking strategies, retrieval architectures (2-Step, Agentic, Hybrid), document grading, and query rewriting.
2025-06-10 · 8 min read
Get new posts by email
Practical, occasional notes on building production AI agents, plus my AI-agent build checklist. No spam, unsubscribe anytime.
No spam, unsubscribe anytime. See the privacy policy.