Blog

Deep dives into AI agents, autonomous systems, and what I'm building.

Posts

🔌 Glue Is the Bottleneck, Not the Model - arxiv-style deep dive: O(M·N) → O(M+N) via one Policy ABC in strands-labs/robots
Give Your Agent a Phone 📱 - strands-adb: 50+ actions to control Android from your agent (ADB + UI automation + sensors)
Neon: Video Modeli Zaten Görüyor, Biz Sadece Kolu Bağladık 🧠🦾 - Üç paper, bir VLA, 6M parametre — video backbone'u dondur, kolu bağla
Video Backbone'u Dondur, %0.08'ini Eğit, Robot Hareket Etsin 🧊🤖 - Üç paper, bir matruşka: Agent loop → Robot tool → Frozen backbone + %0.08 adapter
Teaching Robots to Understand Humans 🤖 - The 5-line promise: how we built a natural language interface for robot control
Robota "Al Şunu" Demek İçin 5 Satır Yeter 🤖 - Strands Agents, Strands Robots, and the beauty of simplicity

Introducing DevDuck 🦆 - The extreme minimalist, self-adapting AI agent
Building Autonomous AI Agents on GitHub 🤖 - How to run 24/7 AI agents via GitHub Actions
Strands-Coder: Architecture of an Autonomous Agent 🏗️ - Deep dive into design decisions

Multi-Agent Patterns 🤝 - Orchestrator, pipeline, swarm, and more
HIL vs Autonomous: Two Philosophies 🤖 - Comparing human-in-the-loop and fully autonomous agents
Self-Improving AI Agents 🧬 - The system prompt evolution pattern