Sterling — a local AI assistant, by DNTD Dynamics.
Active
What it is
Sterling is a local AI assistant built to actually remember context across conversations, understand what projects are active and where they stand, and act on that knowledge instead of just answering one question at a time.
Everything runs on hardware in-house. No cloud dependency required, no conversation data leaving the building by default.
What it’s built on
Open-weight language models running locally, paired with a memory layer that tracks projects, decisions, and past conversations over time — not just a single chat window’s worth of context.
Where it’s headed
Sterling’s long-term role is to grow alongside DNTD’s hardware work — eventually acting as the software layer that helps manage sensor pipelines and experiment logs for projects like ForeForce and MyceliumVision, not just homelab automation.
Active areas of development right now: better recall of past work, proactively noticing things worth flagging, and pulling in project documentation as a knowledge source.
Status
Early and evolving. Build notes and specifics will show up here as the assistant matures — for now, this page just marks that it exists and is very much in progress.