Sabbir Hossain Ujjal

"Hello World!"

I’m a Ph.D. student in ECE Department at Stevens Institute of Technology under the supervison of Prof. K.P (Suba) Subbalakshmi. My research focuses on Natural Language Processing (NLP), Multimodal Modeling, Audio/Speech Processing, and Trustworthy AI.

Previously, I worked as a Machine Learning Engineer (Team Lead) in Advanced Chemical Industries (ACI) Limited, where I collaborated with a dynamic team to build AI-driven solutions that optimized business operations. I focused on adapting NLP and speech models for Bangla, tackling the challenges of building AI systems for low-resource languages.

Before that, I obtained my B.Sc. degree in Electrical and Electronic Engineering (EEE) from Bangladesh University of Engineering and Technology (BUET), where I completed my thesis on NLP and Conversational AI under the supervision of Professor Dr. Mohammad Ariful Haque.

News

Sep 3, 2025	Started my Phd in Computer Engineering at Stevens
Aug 15, 2025	Resigned from ACI ML team. I will soon start my PhD journey at Stevens.
Aug 15, 2024	Promoted to Team Lead of Advanced Chemical Industries (ACI) Machine Learning Engineer Team.
Aug 9, 2024	My team ServerDown secured 2nd runner up place on `IEEE MIPR 2024` workshop task. We presented our technical report of our solution approach in the conference. `[Conference Presentation]`
Jun 3, 2024	My team ACI_ServerDown has become the champion of the biggest data analysis event in Bangladesh Robi Datathon 3.0 organized by Robi Axiata Limited, outshining 1,000 teams formed by 3,500 talented individuals. `[Presentation]`

Selected Publications

mTOVA: A Multilingual Task Oriented Virtual Assistant for Human Computer Communication

Sabbir Hossain Ujjal, A F M Mahfuzul Kabir, and Mohammad Ariful Haque

2023 IEEE International Conference on Telecommunications and Photonics (ICTP), 2023

Abs

A task-oriented virtual assistant(VA) refers to an artificial intelligence driven system that can assist users to perform daily activities. The utilisation of deep learning algorithms has enabled VAs to attain noteworthy advancements in high-resource languages, such as English. However, languages with limited resources, such as Bengali, have not encountered significant advancements in this domain. In this paper, we propose a multilingual task oriented voice-to-voice conversational agent, which is capable of proficiently managing diverse tasks such as weather forecast, date & time query, hospital and blood bank search etc in both Bengali and English language. Our developed system can understand voice command using Automatic Speech Recognition (ASR) and Natural Language Understanding Unit (NLU). Then the system generates an appropriate reply by gathering information from the internet via APIs, data retrieval techniques and by employing dialogue management. Finally, Natural Language Generator (NLG) and Text to Speech (TTS) techniques are used to construct and deliver proper response. We integrated all the units using RASA framework and python script. Our developed ASR system has an average word error rate of 13% and NLU system has an intent and entity extraction accuracy of 93% and 96.2% respectively. The overall action prediction accuracy of our developed system is 99.4%.