#architecture

Articles tagged with #architecture

Your AI Agent Has 12 Hardcoded API Keys and You Call It "Production-Ready" — Dependency Injection Patterns for Agent Systems
Hardcoded LLM clients, inline API keys, baked-in model names — your agent works until you need to change anything. Here are 5 DI patterns that make agents testable, swappable, and production-ready.
Mar 22, 202611 min read2
Your Multi-Agent System Works Until Two Agents Touch the Same File
6 coordination patterns that prevent agents from overwriting each other, losing context, and retrying forever
Mar 14, 202611 min read
Your AI Agent Just Lost 3 Hours of Work. Here's Why.
Building crash-resilient agents with checkpoint recovery patterns from database systems
Mar 7, 202612 min read1
The Supervisor Pattern: When Your AI Agent Needs a Manager
Most AI agents fail silently. They hallucinate. They loop forever. They make a single API call fail and give up entirely. If you've built even one agent, you know the pain. The problem isn't the LLM. It's the architecture. This is where the Superviso...
Feb 25, 20264 min read2

#architecture - Netanel Systems by klement Gunndu