BUILDwith REMEINIUM

@siyabasa/singlish

docs

Singlish-to-Sinhala transliteration

UgannA_Siyabasa

product

Sinhala FastText Model

Separate before you Compress

artifact

WWHO: A Tokenizer, that feels the language

Here we GO

Explore our departments. Each one is a focused domain of engineering and research.

⬡

AI

We believe AI should feel less like machinery and more like magic. At Remeinium AI we craft intelligent systems that don't just process data—they understand context, nuance, and intent. — Intelligence for a greater tomorrow...

◈

Labs

Where experimental engineering happens. We build the tools, libraries, and open-source foundations that power the Next Big Civilization. — Precision built. Community driven.

COMPREHENSIVE

All you need to do is refer the documentation

@siyabasa/singlish

A deterministic, high-performance transliteration engine for Singlish (Romanized Sinhala) with phonetic precision and zero-latency IME support.

Embedding_Siyabasa

Complete API Playground for UgannA_Siyabasa FastText Model

Newsium

A chronicle of our journey in preserving humanity's most meaningful ideas. Explore our latest announcements, technical updates, and stories from the bridge of the next big civilization.

ReleaseDecember 15, 2025

UgannA_Siyabasa V2 Released

Major update to our Sinhala text generation model with improved coherence and instruction following.

ReleaseNovember 20, 2025

@siyabasa/singlish Now on NPM

The Singlish transliteration library is now available as an npm package for easy integration.

ResearchOctober 5, 2025

Sinhala Tokenizer Research Published

Our research on subword tokenization strategies for low-resource Sinhala NLP is now publicly available.

DocsSeptember 18, 2025

Embedding_Siyabasa Documentation Live

Complete API documentation and guides for our Sinhala semantic embedding model.