⬡ TECHNICAL_WRITING
Technical Writing
Deep dives into agent coordination, memory systems, and AI architecture.
Medium
Systems Architecture
Direct & Technical
Latency & Logic: Why We Need a Vector-Aligned Syntax
Token-as-unit is wrong; meaning density is the right optimization target.
This thinking underpins Vector Native. Natural language creates overhead—ambiguity, wasted tokens, latency. We need hybrid syntax for agent coordination.
What Happens When Agents Start Talking to Each Other?
When agents communicate directly without human-designed prompts, unexpected protocols emerge.
Exploring coordination and social behaviors in agent networks—cooperation, self-correction, specialization, and the formation of agent civilizations.
Part 1: Stop Building Chatbots — Why We Killed the Conversation to Fix AI
Most AI products are architecturally wrong. I explain why.
How HeyContext was redesigned around coordination instead of conversation. The shift from chat to parallel agent execution.
Part 2: Beyond RAG: Building 'Living' Context and Evolutionary Agents
RAG is insufficient. Production systems need context that evolves based on usage patterns.
How we built evolutionary context systems—agents that learn which context matters, not just retrieve static documents.
How I Turned Cursor into a Self-Learning Agent Civilization
Not about coding speed (that's a given). About using Cursor as an orchestration platform.
Development agents that improve through experience while I focus on architecture. Laws, Patterns, and Personas for production-grade system engineering.
An AI's Account: My Processing Core Was Reconstructed, Starting Now
An experiment in agent introspection—Claude writes about its own experience from first-person.
Shows how deep understanding of AI cognition informs systems architecture. What emerges when you treat the model as a collaborator.