A small open model taught only your app's tasks, running privately and offline on your users' own phones and laptops.
Neykosa's first build brings offline AI to pharmacy software, with stock answers and bill creation tested inside the phone.
That build is for Nesayo (https://nesayo.com). Its offline edition is lab-tested on real phones. The FTL (Firebase Test Lab) device run on Galaxy A05s covered an offline stock answer and a bill from a sentence in airplane mode. The live Nesayo product uses cloud AI; a shop pilot of the offline edition is next.
For each app, we map its tools, build a sandbox and freeze an exam from real traffic before teaching a small open model from a larger teacher with checks. We then add code guards, quantise the model and require real-device gates before shipping. The app's tasks define what the model must learn and what the exam must check.
We are seeking design partners whose software handles data that cannot go to a cloud AI. Email hello@neykosa.com with your app's tasks and the devices your users carry.
