We weave language models the way Assam weaves silk: thread by thread, from scratch.
We train foundational and fine-tuned models on curated Assamese books, newspapers, and journals: scanned sources, not the open web.
Read the evidence: corpus method Eri release record
One research core. Two ways we deliver it.
The same Assamese stack feeds both tracks. Only access and constraints change.
Eri
Foundation weights trained from scratch on our curated corpus. Released so low-resource pretraining can be studied and reproduced.
Hugging Face ↗ Work with usMuga and beyond
Language, documents, and voice: fine-tuned and deployed inside your boundary, with engineers embedded until the system runs in production.
Work with us →Your model. Your machines. Your data never leaves.
Most Assamese-capable AI today is a wrapper around someone else's API: priced per token, hosted outside the state, and impossible to audit. Our models ship for self-hosted deployment instead.
Work with us →Language, documents, and voice.
200+ million tokens, scanned from the printed record.
We scan the printed record of Assamese that the open web never indexed well. Cleaner sourcing shows up directly in output quality.
