Lecturer · Dept. of CSE, AUST
Aspiring Ph.D. student in Natural Language Processing, focused on making NLP systems more robust, trustworthy, and reliable for real-world use.
Get to know me
I completed my B.Sc. in Computer Science and Engineering (CSE) at Ahsanullah University of Science and Technology (AUST) in January 2025, graduating first in my department. Through active engagement in research and academic activities, I have developed a strong interest in advancing the field of Computer Science while continuing to grow as a researcher and educator.
Teaching has taught me as much as it has let me teach. Explaining complex concepts and engaging with students' questions has continually challenged me to deepen my own understanding and think beyond the classroom. These experiences have motivated me to pursue a Ph.D., where I hope to receive rigorous research training, collaborate with leading researchers, and investigate challenging problems in Natural Language Processing with real-world impact.
My research interests lie in Natural Language Processing, Large Language Models, and the trustworthiness and reliability of NLP systems, particularly for reasoning-heavy tasks such as multi-step mathematical problem solving. I am currently seeking Ph.D. opportunities where I can continue this line of research.
Research output
Presents the "PatiGonit" dataset of 10,000 Bengali math word problems and fine-tunes transformer-based models (Basic Transformer, mT5, BanglaT5, mBART50) to translate word problems into equations. The mT5 model achieved the highest accuracy at 97.30%.
Introduces SOMADHAN, a dataset of 8,792 complex Bengali math word problems with manually written, step-by-step solutions, and evaluates GPT-4o, GPT-3.5 Turbo, LLaMA, DeepSeek, and Qwen models using zero-shot and few-shot Chain-of-Thought prompting. LLaMA-3.3 70B achieved the highest accuracy (88%) with few-shot CoT prompting.
Preprint (arXiv)Beyond the classroom
A Public Trust under the Ministry of Water Resources, Government of Bangladesh.
Academic record
CGPA: 3.917 / 4.00 — Graduated in the first position in the CSE department.
GPA: 5.00 / 5.00
GPA: 5.00 / 5.00
In the classroom
Courses taught, by semester:
What I've built
Projects with published source code. See GitHub for the rest of my work.
Optimizes blood donation and inventory management with an efficient, life-saving approach.
An adventurous shooting game filled with action, thrill, and epic battles.
An Android app that helps users discover their ideal home through a user-friendly rental house search experience.
Provides real-time alerts and crucial data to help mitigate flood risks.
Toolbox
Let's connect
I'm happy to connect about research collaboration or Ph.D. opportunities.