What I'm building, running, and thinking right now.

Agent Infrastructure Platform

Building the operational backbone for autonomous agent systems. Mission Control, orchestration, memory OS (Akashic), decision engines. Infrastructure for running agent fleets at scale.

Content Engineering Pipeline

blog-autopilot → jeremyknox.ai Academy. 226 lessons across 16 tracks. YouTube → transcript → Claude article → Leonardo image → GitHub PR → live. Zero manual steps. Distribution engine for knowledge products.

InDecision Live Trading

Six-factor bias model trading live across Polymarket, Hermes (prediction markets), and crypto derivatives. Continuous optimization: conviction scaling, fee efficiency, regime detection. Real capital at risk.

Tesseract Intelligence

Competitive intelligence system. Real-time signal aggregation, pattern detection, strategic doctrine. Foundation for decision engines across all platforms.

54 apps, 40+ crons

Persistent AI agent orchestration on Mac Mini. Routes tasks, spawns Claude Code sessions, manages async workflows. Nerve center for the entire system.

Multi-market live

Trading live on Polymarket (binary options), Hermes (prediction markets), and crypto derivatives. Six-factor model with conviction scaling, fee optimization, regime detection.

226 lessons live

Blog-autopilot generating 32+ articles. Academy lessons streaming to jeremyknox.ai. YouTube → transcript → article → image → PR → production. Fully automated.

50+ monitors

System health and alerting. Watches trading bots, content pipelines, infrastructure. Gemini Flash post-mortems on anomalies.

Agent Infrastructure R&D

Building systems for autonomous agents to operate at scale. 325+ concurrent agents. Memory systems that learn and compound. Decision engines that scale with conviction. Real-time optimization across multiple problem domains.

AI Engineering Research

Applied research on agentic systems in production. Self-improvement loops. Orchestration at scale. How to build systems that make decisions under uncertainty with real consequences (trading capital, deploying code, managing infrastructure).

On Intelligence

by Jeff Hawkins

The End of the World Is Just the Beginning

by Peter Zeihan

Agent autonomy vs. control. How much judgment do we encode? How much do we reserve for the operator? The scaling laws for oversight.

Memory and compounding. Every mistake becomes a rule. Every rule prevents the next mistake. How does compound learning work at fleet scale?

The alignment problem as operational, not philosophical. We're building systems that make real-world decisions (trading capital, deploying code, managing infrastructure). Correctness isn't optional.