Education

Doctor of Philosophy (Ph.D.)University of California, Riverside
Computer Science and EngineeringSeptember 2022 – June 2026
Research Topics: Multimodal Machine Learning, Computer Vision, and NLP
Master of Science (M.Sc.)University of California, Riverside
Computer Science and EngineeringSeptember 2022 – December 2024
Research Topics: Multimodal Machine Learning, Data Mining
Bachelor of Science (with Honors)Bangladesh University of Engineering & Technology
Computer Science and EngineeringFebruary 2015 – April 2019
Research Topic: Natural Language Processing

Work Experience

Applied Scientist (L5)June 2026 – Present
Amazon Lab126Sunnyvale, California, USA
Topics: Multimodal Machine Learning, Computer Vision, Generative AI
  • Design and fine-tune multimodal machine learning models to enhance robustness and generalization in real-world applications.
  • Build scalable data preprocessing and evaluation pipelines to enable efficient model training and validation.
  • Conduct research, experiments, and technical writing to support publications and advance applied AI initiatives.
Graduate Student ResearcherSeptember 2022 – June 2026
University of California, RiversideCalifornia, USA
  • Worked on the Modality Self-Play (MSP) framework to enable zero-shot multimodal composition in MLLMs without paired cross-modal, non-text training data.
  • Proposed a parameter-efficient adaptation technique that enhances the missing-modality robustness of multimodal models and is applicable to diverse modality combinations, datasets, and tasks.
  • Designed a novel fusion strategy and model, MMSFormer, for multimodal material and semantic segmentation that outperforms state-of-the-art models across all modality combinations.
  • Proposed Masked Modality Projection (MMP) which handles missing modalities by learning to project available inputs, improving robustness across various missing modality scenarios.
  • Designed Cross-Modal Proxy Tokens (CMPTs) with low-rank adapters to enhance missing modality robustness of multimodal models, outperforming existing baselines across five datasets.
  • Developed a graph-based approach to model and analyze social media user interactions, revealing key insights into community dynamics, user behavior, and content preferences.
Teaching AssistantSeptember 2025 – December 2025
University of California, RiversideCalifornia, USA
Course: CS 224/EE 242A: Foundations of Machine Learning
Applied Scientist Intern (L5)January 2025 – June 2025
Amazon Lab126Sunnyvale, California, USA
Topics: Multimodal Machine Learning, Generative AI
  • Designed and fine-tuned multimodal machine learning models to enhance robustness and generalization in real-world applications.
  • Built scalable data preprocessing and evaluation pipelines to enable efficient model training and validation.
  • Conducted research, experiments, and technical writing to support publications and advance applied AI initiatives.
Teaching AssistantSeptember 2024 – December 2024
University of California, RiversideCalifornia, USA
Course: CS 171/EE 142: Introduction to Machine Learning and Data Mining
Teaching AssistantSeptember 2023 – December 2023
University of California, RiversideCalifornia, USA
Course: CS 171/EE 142: Introduction to Machine Learning and Data Mining
Senior Engineer, Software Research and EngineeringJuly 2019 – August 2022
bKash LimitedDhaka, Bangladesh
At bKash:
  • Developed a solution for receiving remittance to bKash digital wallet directly from platforms like Payoneer, Wise, and Remitly, resulting in a 17% increase in the total number of active users.
  • Worked with bKash Customer App team to link the bKash digital wallet to 36 government and private banks for real time and secure money transfers.
  • Worked with bKash Merchant App team to build a real-time dashboard for small to large businesses.
  • Collaborated with bKash Payment Gateway team to implement a fast and secure payment solution, resulting in a 23.2% increase in user engagement and 12.6% increase in total yearly revenue.
  • Also implemented send money, mobile recharge, cash out, and deposit pension scheme (DPS) features.
Engineer, Mobile ApplicationMay 2019 – July 2019
Samsung R&D Institute BangladeshDhaka, Bangladesh
At Samsung:
  • Worked as an iOS application developer in Mobile Application Group 1.
  • Primary responsibilities included developing mobile apps, ensuring code quality, and fixing critical bugs.

Skills

  • Libraries/Frameworks: PyTorch, TensorFlow, Hugging Face, Scikit-learn, OpenCV, NumPy, Pandas, Matplotlib, Flutter, Django
  • Programming Languages: Python, Dart, Java, C/C++
  • Databases: MySQL, PostgreSQL, HBase, Firebase Firestore, Firebase Realtime DB
  • Software/Technologies: Anaconda, PyCharm, Android Studio, VSCode, Firebase, REST API, GraphQL, FastAPI, Git
  • Soft Skills: Communication, Collaboration, Problem Solving

Services

Reviewer2026
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Reviewer2025
The 39th Annual Conference on Neural Information Processing Systems (NeurIPS)
Reviewer2025
The 59th Annual Asilomar Conference on Signals, Systems, and Computers (ASILOMAR)
Reviewer2025
International Conference on Computer Vision (ICCV)
Reviewer2025
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Reviewer2025
IEEE Transactions on Computational Imaging (IEEE TCI)
Reviewer2024
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Leadership

Vice PresidentAugust 2018 – May 2019
Greater Rangpur Students' Welfare Association, BUET
Class RepresentativeApril 2016 – April 2019
Department of Computer Science & Engineering, BUET