About Me
I am an AI Researcher and Part-time Faculty member in the Department of Computer Science & Engineering at United International University (UIU). Previously, I worked as a Software Engineer in the core Machine Learning team at Apurba Technologies Ltd., where I focused on building scalable ML services and VLM-based OCR systems.
My primary research interests lie at the intersection of Multi-modal Learning and Multimodal AI Safety. Specifically, I am interested in:
- Multi-modal Representation Learning: Aligning representations across vision, text, and audio.
- Multi-modal AI Safety & Alignment: Evaluating and defending against adversarial threats on video/multimodal media.
- VLM-based Document Processing: Developing advanced architectures for Bengali Optical Character Recognition and low-resource LLM pipelines.
Note: You can also trigger my printer-friendly CV instantly by pressing Ctrl + P or Cmd + P anywhere on this site.
Education
B.Sc. in Computer Science and Engineering
United International University, Dhaka, BangladeshCGPA: 3.77/4.00 (Distinction: Cum Laude) | Major CGPA: 4.00/4.00 (Data Science)
Thesis: AVLoS: Audio-Visual Long Text Scene Summarization
Supervisor: Prof. Dr. Swakkhar Shatabda
Publications
An audio video-based multi-modal fusion approach for speech emotion recognition
Knowledge-Based Systems (Elsevier), 2026.
BornoDrishti: Leveraging Vision Encoders and Domain-Adaptive Learning for Bangla OCR on Diverse Documents
In the 19th Conference of the European Chapter of the Association for Computational Linguistics (EACL) Industry Track. ACL, 2026.
Beyond Characters: Position-Aware Metric and a Lightweight LLM for Low-Resource Bangla Punctuation Restoration
In 2025 IEEE International Conference on Big Data (BigData). IEEE, 2025.
IPBlocks: A Blockchain Ecosystem for Secure IP Registration and Decentralized Marketplace
In TENCON 2025 - 2025 IEEE Region 10 Conference (TENCON). IEEE, 2025.
E-MedViTR: Enhanced Vision Transformers with Registers for Biomedical Image Classification
In 2024 6th International Conference on Electrical Engineering and Information & Communication Technology (ICEEICT), pp. 664-669. IEEE, 2024.
Watch, Listen, Understand, Mislead: Tri-modal Adversarial Attacks on Short Videos for Content Appropriateness Evaluation
Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops, pp. 2955-2964, 2025.
MIMIC: Multimodal Islamophobic Meme Identification and Classification
3rd Muslims in ML Workshop – NeurIPS, 2024.
Experience
Part-time Faculty
United International University- Teaching core applied computer science courses, focusing on algorithms and software engineering.
- Collaborating with the department of CSE to grade, provide constructive feedback, evaluate student performances, and provide industry insights to enrich course outlines.
- Coordinating with Teaching Assistants (TAs) to create unified course materials and lab outlines.
Software Engineer
Apurba Technologies Ltd.- Engineered scalable, robust, and optimized core ML inference services, resulting in a 10% performance speedup.
- Co-created the first CLIP-style word recognizer for VLM-based Bengali OCR, achieving a word-level accuracy of 95%+.
- Collaborated within a cross-functional ML research team to complete, document, and present critical project deliverables 2-3 weeks ahead of scheduled deadlines.
Lecturer (Contractual)
United International University- Delivered lectures and conducted laboratory sessions for core Computer Science courses.
- Designed and graded examinations, assignments, and projects to evaluate performance and provide feedback, improving students' average lab performance by 7%.
- Courses Conducted: Object-Oriented Programming Lab, Data Structures and Algorithms 1 Lab, and Database Management Systems Theory.
Undergraduate Assistant
United International University- Facilitated course content delivery and prepared hands-on laboratory exercises in coordination with department faculties.
- Evaluated assessments, supervised weekly lab submissions, and served as a panel judge for student projects.
- Authored and structured the country's first-ever undergraduate course materials for the newly introduced Data Science program.
- Courses Supervised: Programming for Data Science, and Object-Oriented Programming for Data Science.
Recent News
📚 Our journal paper "An audio video-based multi-modal fusion approach for speech emotion recognition" has been accepted for publication in Knowledge-Based Systems (Elsevier)! (Q1 Journal, Impact Factor: 7.6).
🎉 Our paper "BornoDrishti: Leveraging Vision Encoders and Domain-Adaptive Learning for Bangla OCR on Diverse Documents" has been accepted at the 19th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2026) Industry Track!
🚀 Our paper on Bangla punctuation restoration "Beyond Characters: Position-Aware Metric and a Lightweight LLM for Low-Resource Bangla Punctuation Restoration" has been accepted at IEEE BigData 2025!
💼 Pleased to announce that I have joined United International University (UIU) as a Part-time Faculty member, teaching core Computer Science and Software Engineering courses!
🏆 Honored to receive the "Best Performing Engineer" award for Project SD-12 from the leadership team at Apurba Technologies Ltd.
🥇 We won the Harvard Health Systems Innovation Lab Hackathon 2025 (Dhaka Hub) for our project HAVEN, a BCI-powered speech-to-speech emotional agent!
🔥 Our paper "MIMIC: Multimodal Islamophobic Meme Identification and Classification" was accepted and presented at the 3rd Muslims in ML Workshop at NeurIPS 2024 in Vancouver, Canada!
✈️ Travelled to Boston, USA to participate as a global finalist in the Hult Prize Summit 2024 representing UIU!
🏆 Won Champion at the Bhashamul: Bengali Regional Text to IPA National NLP Datathon! Our team solved Bengali regional IPA transcription.
🌍 Travelled to Switzerland and won the prestigious Gold Award (World Champion) at the International Blockchain Olympiad (IBCOL 2023) in Lauterbrunnen!
Projects
Research Projects
A novel Vision-Language model (VLM) framework and domain-adaptive learning strategy for optical character recognition of diverse, complex Bengali documents. Associated with Apurba Technologies Ltd.
Solving regional IPA transcription in Bangla using Dialect Guided Tokens (DGT). Won 1st place in the Bhashamul National NLP Datathon.
Engineering & Hackathon Projects
Real-time AI-based caretaking mobile app simplifying aged care through caretakers. Full-stack development including backend and AI engine.
Brain-Computer Interface (BCI) powered mental health speech-to-speech conversational agent recognizing and responding with emotions. Won Harvard HSIL Hackathon Dhaka Hub.
Web3 decentralized app for Intellectual Property protection, licensing, and marketplace. Led backend, frontend, and smart contract development. Won Gold at IBCOL 2023.
Achievements & Awards
Gold Award (World Champion)
International Blockchain Olympiad (IBCOL 2023)
2023Winner (Dhaka Hub)
Harvard Health Systems Innovation Lab Hackathon
2025Best Performing Engineer
Project SD-12, Apurba Technologies Ltd.
Oct 2025National Champion
Bhashamul: Bengali Regional Text to IPA National NLP Datathon
2024Academic Excellence Scholarship
United International University
2020 — 2024Global Summit Participant
Hult Prize Global Summit (Boston, USA)
2024Champion
Hult Prize OnCampus Round (UIU)
2024Silver Award (1st Runner Up)
Bangladesh Blockchain Olympiad (BCOLBD 2023)
20231st Runner Up
Intra-University Deep Learning Sprint
2023National Finalist
ICT Innovation Grant (for NurtureAid)
2023Skills
Programming Languages
Machine Learning & Research
Frameworks & Databases
DevOps & Tools
Extracurriculars & Media
-
UIU Computer Club (UIUCC) General Secretary (2022 — 2024)
Led club activities, organized multiple national hackathons, managed a committee of 50+ members, and directed workshops for 1000+ students.
-
National TV Feature (Jamuna TV) Watch Interview
Interviewed by Jamuna TV on our team's triumph at IBCOL 2023 in Switzerland and the potential of blockchain technology in Bangladesh.
-
Automatic Speech Recognition (ASR) Workshop Co-Trainer (Collaborative Event)
Conducted a hands-on workshop on ASR models hosted by Bengali.AI and the Islamic University of Technology (IUT) Computer Society.
-
dApp Development Workshop Lead Trainer
Led a hands-on workshop on building decentralized applications (dApps) from scratch for the students of United International University.