Orature AI - Pioneering Urdu Language AI

Mission: Orature AI is dedicated to advancing the frontiers of Artificial Intelligence, and Language Models for the Urdu language. We aim to develop computationally efficient, culturally-aware, and accessible language technologies that empower local communities, researchers, and businesses. Our work focuses on bridging the linguistic digital divide and promoting equitable and sustainable AI development.

Vision: To be a leading force in creating and democratizing state-of-the-art NLP resources for Urdu, a low-resource language, fostering innovation and inclusivity in the global AI landscape.

About Us

Orature AI has emerged from the foundational work of the ALIF الف project, a Final Year Project at Habib University (Spring 2025). Our core team comprises passionate researchers and engineers committed to open-source principles and collaborative innovation.

Core Team (Founders of ALIF الف):

• Syed Muhammad Ali Naqvi
• Zainab Haider
• Syeda Haya Fatima
• Hammad Sajid
• Ali Muhammad Asad

Supervisor of ALIF الف:

• Dr. Abdul Samad
• Dr Inayat Ullah

Affiliation:

• Habib University, Dhanani School of Science and Engineering

Our Focus Areas

• Data Curation & Tokenization: Novel creation and meticulous preprocessing of large-scale, culturally relevant datasets and language-specific tokenizer.
• Urdu Language Model Development: Creating robust pretrained and instruction-tuned Small Language Models (SLMs) for Urdu.
• Low-Resource NLP: Developing scalable frameworks and methodologies for building language models for underrepresented languages.
• Open Source Contribution: Sharing models, datasets, and research findings with the global community.
• Sustainable AI: Advocating for efficient and environmentally conscious AI practices.

Our Flagship Project: ALIF الف

The ALIF الف project represents our initial and core contribution, featuring a series of Urdu pretrained generative models, custom tokenizers, and comprehensive datasets.

Get Involved

• Explore our Models & Datasets: Browse our contributions on the Hugging Face Hub.