AI with a life
beyond the chat box.

I build companions, operators, and tools.
These are four of them.

Explore the work ↓

THE SUPPORTING WORK

Under the hood.

The memory, routing, and engineering tools developed alongside the products.

MuninnMemory & evidence.

My shared memory substrate: durable context, scoped retrieval, and a record of where remembered information came from.

MímirContext for coding agents.

A persistent model of a codebase, built to recover relevant context and architectural boundaries for a specific task.

Local AI LabInference & routing.

The serving, routing, and operational work behind running these applications on local AI infrastructure.

More engineering & earlier work ↗

THE PERSON BEHIND THE PROJECTS

I want to know what happens when you actually build it.

I’m Auston. My background spans healthcare, operations, and hands-on AI engineering. Healthcare and operations keep me thinking about the person using a system: what they need to accomplish, what can go wrong, and whether the thing we built actually helps.

Curiosity drives the range of projects here. I want to know whether an AI can maintain continuity, work across my tools, or become useful inside a specialized domain. Building gives those questions somewhere concrete to go—and usually reveals harder questions underneath.

I like working through the whole problem: shaping the experience, connecting the parts, running the infrastructure, and testing the assumptions. Sometimes that means building a product. Sometimes it means building the machinery it needs. Sometimes an experiment tells me to keep the simpler approach.

That combination of curiosity and practical responsibility is what I bring to a team. If you’re working on a difficult problem and want someone who will get deeply involved in making it work, I’d like to talk.