Sheffield, United Kingdom
Research Engineer · Data Scientist · CTO
A research engineer with over a decade of experience in software engineering, data engineering, and data science. Specialized in architecting scalable microservices and leading technical teams. Currently leading R&D at Penta Global.
01
Head of R&D
Applied research, NLP and computer vision
Data Consultant
Trade-based money laundering research with Bangladesh Bank
With over 12 years of industry experience and 5 years in research and teaching, I specialize in building scalable, robust, and efficient solutions. My expertise spans software engineering, data engineering, and data science, with a proven track record of leading technical teams and delivering impactful projects.
I've built highly scalable applications handling 8000+ RPS, developed data fusion platforms used by 18 law enforcement agencies, and conducted research in NLP and computer vision. My work bridges academic research and industry applications, focusing on real-world impact.
Currently I'm leading applied research at Penta Global and research on trade-based money laundering with Bangladesh Bank.
02
Courses I build while learning something. Each one lives in its own repo and is published here.
Learning Japanese from zero — hiragana and katakana, then grammar and reading — with Arabic comprehension running alongside as a second track.
/learning/japanese →
Spoken-first companion to Al-Arabiyyah Bayna Yadayk Book 1: lessons, audio dialogue readers, flashcards, and sentence-builder practice.
/learning/arabic →
A hands-on Docker course taught from the terminal up — first the Linux floor every container stands on, then Docker itself. Labs run inside a disposable Ubuntu container.
/learning/docker →
04
05
2025
A digital repository documenting the July-August 2024 mass uprising in Bangladesh. Features an interactive map of verified deaths, a detailed timeline of events, and preserves digital evidence of human rights violations.
2026
06
March 2024 - June 2024
2024 - 2026
Advanced security platform.
2023 - 2025
Responsible for industrial products which increased collaboration between academia research and industry. Managed industry standard training programs.
September 2021 - March 2023
The Coding Crowd
April 2020 - September 2021
October 2019 - February 2020
May 2019 - April 2020
January 2018 - 2020
October 2017 - 2021
Worked with a talented cybersecurity group in Bangladesh, with a wide interest in cybersecurity.
January 2017 - June 2018
January 2015 - April 2019
October 2013 - September 2014
Responsible for both Android and iOS mobile platforms.
May 2012 - September 2013
Cross-platform game development and cross-platform mobile application development for both Android and iOS platforms.
07
5-DIALECTS-BN: Unmasking the Impact of Transliteration on Bangla Dialectal LLMs
EMNLP 2026 Main Conference (Core Rank: A*)Accepted
BanglaProtha: Evaluating Vision Language Models in Underrepresented Long-tail Cultural Contexts
IEEE/CVF WACV 2026 (Core Rank: A)Published
Evaluating Large Vision Language Models on Bangla Medical Visual Question Answering
Findings of the ACL: ACL 2026 (Core Rank: A*)Published
How Good LLMs Are at Answering Bangla Medical Visual Questions? Dataset and Benchmarking
AAAI 2026 Bridge Program on AI for Medicine and Healthcare (PMLR 317)Published
Decoupling Representation for Fine-tuning with Linear Probing and LoRA on Robust Downstream Adaptation
Venue TBDUnder Review
CEKALA: CKA-Guided Layer Selection for Efficient Vision–Language Adaptation
Venue TBDUnder Review
BanTH: A Multi-label Hate Speech Detection Dataset for Transliterated Bangla
Findings of the ACL: NAACL 2025 (Core Rank: A*/A)Published
ChitroJera: A Regionally Relevant Visual Question Answering Dataset for Bangla
ECML-PKDD 2025 (Core Rank: A)Published
LP-FT-LoRA: A Three-Stage PEFT Framework for Efficient Domain Adaptation in Bangla NLP Tasks
Second Workshop on Bangla Language Processing (BLP-2025)Published
Robustness of LLMs to Transliteration Perturbations in Bangla
Second Workshop on Bangla Language Processing (BLP-2025)Published
Benchmarking Large Language Models on Bangla Dialect Translation and Dialectal Sentiment Analysis
Second Workshop on Bangla Language Processing (BLP-2025)Published
PentaML at BLP-2025 Task 1: Linear Probing of Pre-trained Transformer-based Models for Bangla Hate Speech Detection
Second Workshop on Bangla Language Processing (BLP-2025)Published
BanglaTLit: A Benchmark Dataset for Back-Transliteration of Romanized Bangla
Findings of the ACL: EMNLP 2024 (Core Rank: A*)Published
Penta ML at EXIST 2024: Tagging Sexism in Online Multimodal Content With Attention-enhanced Modal Context
EXIST Lab at CLEF 2024Published
Penta-nlp at EXIST 2024 Task 1-3: Sexism Identification, Source Intention, Sexism Categorization in Tweets
EXIST Lab at CLEF 2024Published
Tools and Techniques Adapted for Teaching Software Engineering Topics Remotely during the COVID-19 Pandemic
HICSS 2022Published
Scrutiny of Electricity Billing and Supply Data as a Probable Proxy for Economic Activities: A Comprehensive Analysis of Power Consumption of Dhaka, Bangladesh
SSRNPublished
08
Computer Science & Engineering
2012
University of Dhaka, Bangladesh