Ayman Mansour

ASR/NLP Researcher, Web Developer, Apps Developer


About Me

Hi 👋🏾

I'm Ayman. an Independant Researcher, I have obtained my MSc degree in Computer Science from Sudan University. My thesis focuses on leveraging Deep learning in supporting Low-resourced Language in the field of Automatic Speech Recognition (ASR). My research interests include appling machine learning solutions for solving real world NLP problems and seeking to improve machine learning models for natural language processing for under-resourced languages.


Machine Learning/Deep Learning.

Automatic Speech Recognition (ASR)

Natural language processing (NLP).

Transfer Learning.

Education & Expericence

Education & Expericence


Master in Computer Science
Sudan University of science and technology
Khartoum - Sudan | 2017 - 2020

After completing the degree curriculums in security track, pursuit machine learning track (self-taught), and intensive research in the field of ASR and NLP for Arabic language and its dialects.

Bachelor of Science in Computer Science

SHARQ ELNEEL University College
Khartoum North - Sudan | 2011 - 2016

Bachelor of Science (honors) in Computer Science .


IT Manager

POLO Frenzy trading company
Khartoum - Sudan | Mar 2019 - Mar 2022

• Defining the development requirements based on the company's needs and setting the necessary plans for the required development units.
• Establishing work plans and priorities for IT management.
• Examining and analyzing the operating problems of the used systems.
• Directing sales supervisor and monitors results of implementation.

Apps Developer & Designer

Tech2Go for smart solutions
Khartoum - Sudan | Mar 2018 - Dec 2018

• Developed Android applications and their backend interface.
• Used google services (Maps, Gson) to make applications more robust and reliable.

Full-Stack Developer

OBAAA for design and programming
Khartoum North - Sudan | Jun 2016 - Dec 2016

• Conceptualized and developed web-based content management systems for small schools and institutes.
• Developed a web-based distributed system with user and group-level permissions.
• Created powerful and flexible systems for users of all technical abilities.




Sudanese Dialect Morph Analyzer
Coming Soon.

Fine-tuned OpenAI whisper
Fine-tuned OpenAI whisper-small pre-trained to support the Sudanese dialect and got WER: 56.32.

Fine-tuning XLSR-WAV2VEC2
Fine-tuning XLSR-WAV2VEC2 pre-trained model on the Sudanese dialect.

Sudanese dialect dataset
Constructing a Sudanese dialect dataset for NLP and speech projects.

Founded oook.sd initiative to enrich colloquial content (Sudanese dialect) to include it in future NLP research and the field of linguistics. The initiative aims to collect Sudanese content (speech and text) and enable access to it by researchers.

Sudanese ASR model
Designed Automatic Sudanese dialect recognition model to convert speech to text.

SW Development

To simplify management task and issue invoices.

Android application for ensuring traffic awareness for both drivers and pedestrians.

ndroid application to overcome gas shortages in Sudan between 2018 – 2019.

Android application to enable users to rent their tractors to any customers.

Al-Gafari & Tech2Go
Introduce and offer both companies' services and solutions for interested customers.

Nora system
For managing schools and institutes to track students' progress also organizing courses.




Contact Me

© Ayman-Mansour. All Rights Reserved. Designed by HTML Codex