Richard Amerman

Richard Amerman is an AI engineer doing independent work exploring what genuinely intelligent systems might require: not just better language models, but architectures capable of true comprehension, continuous learning, and building dimensional models of the world they operate in. After 30+ years navigating the intersection of infrastructure, development, and emerging technologies, he has found the work he believes he was always meant to do.

The Arc

His path here was anything but direct. Richard grew up in Sitka, Alaska, a small town on an isolated island in Southeast Alaska. The combination of that remote setting with world-class cultural influences, including a community-built public radio station and an arts camp that drew people from across the country, created a generation of young people who saw the entire world as their playground. Friends scattered to Oxford, Sweden, and points far beyond. That sense of possibility never left.

After high school, Richard joined the US Army and was recruited into the 3rd US Infantry Division, the Old Guard, the Presidential Honor Guard. He taught himself fife, passed the tryout for the Old Guard Fife and Drum Corps, and spent the rest of his service performing at state ceremonies and diplomatic events. After the Army, he taught hang gliding at Kitty Hawk Kites in Nags Head, NC, became an EMT and First Aid/CPR instructor, and continued pursuing his original plan to become a music teacher. He played traditional Irish flute in gigging bands across the Seattle scene throughout the '90s, sang bass with community ensembles, and carried that musical discipline through every career transition.

The plan was always music education. Richard held onto that path well into his thirties, accumulating enough credits for nearly two degrees along the way. But while working a contract QA role in Seattle, the obvious finally registered: he was already deep in technology, he was good at it, and the career had been building itself around him for years. He made the pivot, founded 7 Tech Northwest, and spent the next two decades building across industries: development, infrastructure, cloud architecture, and DevOps. The last twelve years have been at CD Baby (Downtown Music Holdings), where he has served as Senior Cloud Engineer leading cloud infrastructure and AI workflow integration.

The Current Chapter

Today that ability to walk into unfamiliar territory and build a working model of the landscape in real time is focused entirely on AI. Richard is building and maintaining multiple open-source projects that tell a connected story: UnaMentis, a voice AI learning platform designed around the conviction that AI should adapt to human learners, not the reverse; Pocket TTS for iOS, a port of Kyutai's 117M parameter TTS model from Python/PyTorch to Rust/Candle for native on-device inference; OpenClaw, a governed multi-agent coding architecture; Agent Vision Team, the precursor governance system built on Claude Code hooks; and Solution Explorer, a codebase architecture visualization tool.

Beyond the engineering work, Richard has published a knowledge distillation framework that lays out an understanding-based approach to machine intelligence, exploring questions about comprehension, grounding, and structural integration of knowledge. He is active in Portland's AI community, attending events regularly and speaking 2-3 times monthly, and is increasingly sought out as a mentor and resource on voice AI. He also serves as a Planning Commissioner for the City of Ridgefield and mentors the local FIRST Robotics high school team.

Intellectual Honesty

Richard is transparent about what he does not have: he does not hold a PhD, he did not come up through a research lab, and his formal education in machine learning is self-directed and ongoing. What he brings is an unusual combination of systems-level engineering depth, rapid first-principles reasoning, and the kind of intellectual hunger that does not stop at the edges of a job description. He is deeply engaged with the global AI community and committed to contributing at the frontier of this field, wherever that work needs to happen.

Core Experiences