About Us
One engineer. European infrastructure. Enterprise-grade AI.
AI moves the work faster. Human judgment decides what ships.
Blue Note Logic builds in active dialogue with tools like Claude and Codex, but the accountability stays human. We use AI for speed, pressure-testing, drafting, refactoring, and exploration, then review the architecture, code, content, and deployment path before anything reaches a client.
That is the operating method behind Do Better Norge, caveauAI, and the production systems we deploy: assisted by AI, governed by people, and measured by whether the result survives real use.
Started as a Thesis. Built into a Company.
Blue Note Logic began as a thesis project: a practical attempt to design and build a better way to work with dense, high-stakes documents.
What started as one person’s research and engineering effort became a working platform, then a repeatable architecture, and now an incorporated company focused on sovereign AI infrastructure and citation-backed document intelligence.
Founder Dave Gilligan developed the first versions while living in Norway, drawing on a background across risk modelling, telecom transformation, and full-stack technology delivery. The goal was never to launch another generic AI wrapper. It was to create a system that could ingest real source material, return accountable answers, and stay under the owner’s control.
That origin still shapes the company today. Blue Note Logic is lean by design, technically opinionated, and built around European hosting, open models, and practical delivery. The thesis became a platform. The platform became a business.
Talk to Blue Note Logic
Reach out directly if you want to discuss sovereign AI, document intelligence, or a practical implementation path.
Locations
United States incorporated
Sales and operations
The Team
A lean team that builds, ships, and operates every day.
Where We Operate
Blue Note Logic is a United States incorporated company with a general, partner-supported operating model. We choose the footprint around the client, the data boundary, and the level of control the work requires.