Local AI
Language models run on your own hardware through Ollama, with local text-to-speech and speech recognition for NTX-1.
How we build
We use technology where it simplifies decisions, organizes information or turns an overlooked problem into a useful product. The problem comes first.
Language models run on your own hardware through Ollama, with local text-to-speech and speech recognition for NTX-1.
Claude, GPT, Gemini, Grok and local models, used where they make a product genuinely better.
Installable, offline-capable web apps built mobile-first, like Snap2Task and CardCommand Pro.
Geolocation, maps, weather and market data behind GrubScoutUSA and AtAGlanceInfo.
Turning complex information into clear interfaces, from body charts in ChartBodyDTx to cost breakdowns in CostInHours.
Python, pywebview and native Windows integration for apps like NTX-1’s floating Detach mode.
Python and PowerShell bots, browser automation and agent workflows for clients and our own operations.
On-device storage where it matters, WebAuthn biometrics and encryption in our finance tools.
Infrastructure