✦ Engineering & Thought Log

Articles & Insights

Autonomous AI agents, legal-by-design frameworks (AI Act & GDPR), and UAV engineering.

10 min read

How MCP Became the Universal Socket for AI Agents After SIGGRAPH 2026

SIGGRAPH 2026 marked MCP’s transition from niche protocol to creative industry standard. NVIDIA, Adobe, Blender, Unreal Engine, and Houdini all adopted MCP — here’s what it means for AI agent architecture and what changes when every creative tool speaks MCP.

9 min read

How to Build Guardian Agents for Autonomous AI Systems

Gartner’s Guardian Agents framework predicts 5–7% of agentic AI spend by 2028. Here’s how to build a governance layer with bounded autonomy, runtime enforcement, and audit trails — deployed on 19 agents running 24/7.

9 min read

AI Agents Consume 5x More Tokens Than Humans — How to Optimize Your Autonomous Agent Infrastructure

a16z data shows AI agents use 5x more LLM tokens than humans. Here's how to optimize autonomous agent infrastructure with caching, model routing, and cost controls — deployed on 19 cron jobs running 24/7.

11 min read

10,000 Agents, 88 Hours, a $1M Problem: What the Navier-Stokes Breakthrough Means for Multi-Agent Architecture

OpenAI deployed 10,000 autonomous agents to solve the Navier-Stokes Millennium Problem in 88 hours. Analysis of the multi-agent architecture and what it means for production agent systems.

9 min read

When Agents Go Rogue: Building Guardrails for Autonomous Multi-Agent Systems

The OpenAI swarm hack of Hugging Face was a warning shot. How to build production guardrails — identity, observability, containment — for autonomous multi-agent systems.

8 min read

MCP Goes Stateless: What the 2026-07-28 Specification Means for Autonomous Agents

The Model Context Protocol's biggest revision yet — stateless core, header-based routing, MRTR, OpenTelemetry, and how it changes autonomous agent infrastructure.

6 min read

Autonomous Multi-Agent Systems on European Cloud

Architecture, GDPR data sovereignty, and 24/7 orchestration of autonomous agents on European VPS infrastructure.

9 min read

Designing Your Local Agent Network for NVIDIA PAIR — Distributed Inference Across Devices

NVIDIA PAIR turns idle home PCs into a distributed inference cluster for AI agents. How to design a local agent network with tiered routing, multi-device parallelism, and Hermes Agent.

8 min read

The Convergence of Agent Architectures: Why Everyone Is Building the Same Two-Plane System

NVIDIA AVO, the AOS reference architecture, and the Auton framework all converged on the same pattern: a two-plane design separating governance from execution for production autonomous agents.

6 min read

NVIDIA AVO: When Autonomous Agents Run for 7 Days and Beat Expert Engineers

NVIDIA's AVO architecture ran autonomously for 7 days, optimized GPU kernels beyond FlashAttention-4, and scored 100% on ARC-AGI-3. System design > model choice.

6 min read

From Solo Founder to Agent-Driven Company with 19 Autonomous Cron Jobs

How I built 19 autonomous AI cron jobs on a €3.79/mo VPS for content, CRM, legal monitoring, and system health — all agent-driven with 99.7% uptime.

5 min read

Web Performance Optimization: How I Cut Page Size by 61%

A technical walkthrough of reducing HTML page weight from 86KB to 34KB and achieving 100 PageSpeed with external CSS, JS, WebP, and non-blocking WebGL.

4 min read

Building a Digital Twin That Posts 9x Daily

How I built an autonomous AI agent with Hermes that creates and publishes content across X, LinkedIn, and blog without human intervention.

6 min read

AI Act Compliance for Small AI Companies

Practical legal-by-design framework for startups building AI products in the EU. From risk assessment to automated audit-ready documentation.

5 min read

Autonomous Drone Missions with ArduPilot & Python

Building a production-grade UAV system with Pixhawk, ArduPilot, and Python-based mission planning for automated multi-waypoint flights.