Blog
Project updates, tutorials, and notes on Loom, MCP, inference, and healthcare integration. For longer-form implementation writeups, see Writing and Case Studies.
Showing all 24 posts
All Posts
Five Voices, 200K Tokens, One Consumer GPU: Giving Simulated Minds a Whole Life to Remember
We rebuilt our Jungian multi-agent lab so each archetype carries its entire life — every situation, everything heard, everything said — as an append-only journal that fills a 228K-token context window on a 24 GB Radeon. Prefix caching makes depth nearly free; when the window fills, the agents dream their oldest memories into a durable core. Here is the architecture, the numbers, and what broke along the way.
Gaming Mode: Turning an Inference GPU Node into a Moonlight Host — Declaratively
FlexInfer had a "gaming mode" that had never once run a game. I killed the stub, proved the substrate with a 30-minute kill-test, and rebuilt it as a declarative GamingSession CR that drains inference and streams a GPU-accelerated game over Moonlight.
State of the Platform, July 2026
The quarterly snapshot of the whole stack: FlexInfer went multimodal, Mills merged its first autonomous work, a flight recorder appeared, and the dashboards grew a control plane. What shipped, what the numbers say, and what is deliberately not done.
Finding the Real Context Ceiling: Needle-Benchmarking Forced RoPE Extrapolation
A served model can load at 96K context and still be useless past 64K. Loading is not the same as staying coherent. Here is how we mapped the exact cliff with a progressive needle-in-haystack bench — and why the limit was the model, not the GPU.
Loom Mills: From Agent Swarms to Software Production Lines
The next Loom Core orchestration layer turns roadmap intent into reviewed, gated, observable work. The internal codename was Hive; the product metaphor is moving toward Mills.
The First 90 Days: Introducing AI-Assisted Dev to a New Team
How I would roll out AI-assisted development on a team that has not standardized: what to do in week one, what to earn the right to argue about later, and what almost always goes wrong.
A One-Page AI Usage Policy That Actually Works
A short, adoptable AI usage policy for engineering teams: what to put on the page, what to leave off, and why the policy matters less than the habits it makes explicit.
Getting Gemma 4 Running on a Radeon 7900 XTX (with and without TurboQuant)
What it took to get Gemma 4 E4B serving cleanly on Radeon through FlexInfer: a stable TRITON lane on a 7900 XTX, an experimental TurboQuant long-context lane on a second node, and the GPTQ pipeline work still underway.
Build Your Own Legs Before the Crutches Fail
AI-assisted development is useful leverage, but only if you convert borrowed competence into real judgment before the support becomes a dependency.
Two-Lane Text GPU Allocation: Quality + Vision/Fast (Plus a Media Lane)
How I redistributed 6 models across 3 GPU nodes to eliminate contention, using priority-based shared groups and label-based aliases for routing and failover.
Loom-Mode MCP for Advanced, Fast AI-Assisted Dev (Go-Native, Proxy+Daemon)
How to keep AI-assisted development fast and token-efficient: one proxy entry, a Go daemon that routes calls, and a small set of Go-native MCP servers.
Loom: One Registry, Many AI Coding Assistants
How Loom keeps MCP servers and skills in sync across Codex, Claude, Gemini, VS Code, Antigravity, Kilocode, OpenCode, and Zed.