Neykosa builds private, offline AI for your app.
A small open model taught only your app's tasks, running privately and offline on your users' own phones and laptops.
Our first build is the offline edition of Nesayo, lab-tested on real phones.
Who is behind Neykosa?
Team Neykosa, based in Bengaluru, India. Contact us at hello@neykosa.com.
What has Neykosa shown so far?
Nesayo is pharmacy software. The live Nesayo product uses cloud AI; its offline edition runs our model inside the phone.
Lab-tested on real phones in Firebase Test Lab (FTL).
- Test phone FTL
- Galaxy A05s, 4 GB, Snapdragon 680. Android 15, airplane mode, run PASSED
- Stock answer, offline FTL
- 4.25 s. The first turn of that run took 13.4 s (right after a print preview); still open
- Bill from a sentence FTL
- 5.76 s
- Model memory FTL
- 673 MB
These measurements describe the Nesayo offline edition, not your app. See measured results with their layers.
How would we build for your app?
Only the AI your software needs, inside the phone.
- Map the task. Map your app's tools and build a sandbox.
- Keep an exam out of training. Freeze an exam from your app's real traffic.
- Teach and check. Teach a small open model from a larger teacher with checks, then add code guards.
- Test on devices. Quantise the model and pass real-device gates on your users' phones.
This is the process for a partner build. Results on your app would need to be measured.
Where we are today
No paying customer, and no pilot running. A shop pilot is next. We have no partnerships, investors or certifications.
Battery use and a cloud cost comparison are not measured. iPhone and NPU support, and laptops beyond the tested desktop build, are roadmap items.
Bring the task your app needs to handle
Our design-partner offer is for software companies whose app handles data it cannot send to a cloud AI, or that pay per-question cloud bills.
Tell us what your software does, one task you want the model to handle, your target devices and what keeps you from using cloud AI.