About

Mofijul is a Tech Lead and Senior Applied Scientist at AWS GenAI. His research focuses on multi-agent reasoning, post-training and finetuning of multimodal LLMs, synthetic data generation, and code generation. He also draws on insights from human and animal multisensory systems to inform multimodal learning.

Mofijul serves as an Area Chair for ICLR, ICML, NeurIPS, and InterSpeech. He is also the Research Liaison Lead of the Amazon-UCLA Science Hub, a Graduate Research Mentor at Johns Hopkins University, an Associate AWS Speaker, and a member of the Amazon Research Award Review Panel.

Mofijul completed his Ph.D. at the University of Virginia (Thesis: Multimodal and Multitask Representation Learning for Perceiving Human Behavior in Embodied Settings) as an ESE-UVA Bicentennial Fellow (2019-2020), working at the Link Lab with Prof. Tariq Iqbal. Previously, he was a Lecturer at the Department of Computer Science and Engineering, University of Dhaka, where he also earned his M.S. and B.Sc. (Hons) in Computer Science and Engineering.

Outside of research, Mofijul enjoys writing code to background music, participating in hackathons and programming contests, and traveling to new places.

Updates

  • Feb 2026: Promoted to Tech Lead & Senior Applied Scientist at AWS GenAI
  • 2026: Serving as Area Chair for ICLR-2026, ICML-2026, NeurIPS-2026, and InterSpeech-2026
  • 2026: Three papers accepted at ACL-2026 (Main Conference, Industry Track, and Demo) — “Ready to Translate, Not to Represent?”, “DocSplit: A Benchmark for Document Packet Recognition and Splitting” [Benchmark], and “IDP Accelerator: Agentic Document Intelligence”
  • 2026: “Energy-Based Transformers are Scalable Learners and Thinkers” accepted as Oral at ICLR-2026 [ArXiv]
  • 2026: “Embodied Referring Expression Comprehension Through Multimodal Residual Learning” accepted at HRI-2026 [ArXiv]
  • 2026: “MATHMIST: A Parallel Multilingual Benchmark for Mathematical Reasoning” accepted at EACL-2026 [ArXiv]
  • Nov 2025: “DM-Codec: Distilling Multimodal Representations for Speech Tokenization” accepted at EMNLP-2025 [ArXiv] [Code]
  • 2025: Featured AWS Blog: Strands AI Agents for the GenAI IDP Accelerator
  • 2025: Featured in Wired: NFL Storytelling
  • 2025: Led org-wide open-source initiative: Agentic Multimodal Document Processing (GenAI IDP) — contributed millions in ARR
  • 2024: Featured in AWS Blog: Accelerate Intelligent Document Processing with Generative AI
  • 2024: Featured in AWS Blog: Healthcare Document Comprehension (Myriad Genetics)
  • 16 Oct 2024: Our work on code generation is featured in AWS Blog
  • 16 Jan 2024: EQA-MX accepted as a Spotlight at ICLR-2024
  • 06 Nov 2023: Joined AWS Generative AI Innovation Center as Applied Scientist II
  • 17 Jul 2023: Joined the Amazon Alexa AI team as an Applied Scientist II Intern
  • 30 Jun 2023: Ph.D. complete!
  • 10 Feb 2023: Accepted to Google’s Computer Science Research Mentorship Program (CSRMP)
  • 20 Jan 2023: “Representation Learning in Deep RL via Discrete Information Bottleneck” accepted at AISTATS 2023 [Paper]
  • 21 Dec 2022: Received the AAAI-23 Scholarship
  • 05 Dec 2022: Ph.D. thesis accepted to the WACV 2023 Doctoral Consortium (with a travel award)
  • 02 Nov 2022: Received the UVA Engineering Endowed Fellowship (Fall 2022 – Spring 2023)
  • 07 Oct 2022: Received the NeurIPS-2022 Scholar Award
  • 16 Sep 2022: “CAESAR: An Embodied Simulator for Generating Multimodal Referring Expression Datasets” accepted at NeurIPS 2022 (Datasets and Benchmarks Track) [Paper]
  • 11 Sep 2022: Attending Ubicomp-2022 to present our work on M3Sense
  • 28 Mar 2022: “A Vocabulary-Free Multilingual Neural Tokenizer for End-to-End Task Learning” accepted at the ACL Workshop on Representation Learning for NLP 2022
  • 22 Mar 2022: “MAVEN: A Memory-Augmented Recurrent Approach for Multimodal Fusion” accepted at IEEE Transactions on Multimedia [Video]
  • 06 Mar 2022: “M3Sense: Affect-Agnostic Multitask Representation Learning using Multimodal Wearable Sensors” accepted at ACM IMWUT/Ubicomp 2022 [Video]
  • 04 Feb 2022: Received the AAAI Conference Scholarship (2022)
  • 03 Dec 2021: “iSpy a Humorous Robot: Evaluating the Perceptions of Humor Types in a Robot Partner” accepted at the AAAI Spring 2022 Symposium
  • 01 Dec 2021: “MuMu: Cooperative Multitask Learning-based Guided Multimodal Fusion” accepted at AAAI 2022 (Main Track, Oral) [Pre-print] [Video] [Flash Talk]
  • 29 Nov 2021: “Who’s Laughing NAO? Examining Perceptions of Failure in a Humorous Robot Partner” accepted at HRI-2022
  • 04 Oct 2021: Received the UVA Engineering Endowed Fellowship (2021-2022)
  • 03 May 2021: Joining the Amazon Alexa AI team as a Summer-2021 Applied Scientist Intern
  • 02 May 2021: Volunteering at ICLR-2021
  • 29 Jan 2021: “Multi-GAT: A Graphical Attention-based Hierarchical Multimodal Representation Learning Approach for Human Activity Recognition” accepted at IEEE Robotics and Automation Letters (RA-L) [Video Demo]
  • 06-12 Dec 2020: Student Volunteer at NeurIPS-2020
  • 12-18 Jul 2020: Student Volunteer at ICML-2020
  • 30 Jun 2020: My first Ph.D. work, “HAMLET: A Hierarchical Multimodal Attention-based Human Activity Recognition”, accepted at IROS-2020 [Video]
  • 07-10 Mar 2020: Student Volunteer at HRI-2020
  • 26 Aug 2019: Received the ESE-UVA Bicentennial Fellowship (2019-2020)
  • 16 Aug 2019: Moved to Charlottesville, Virginia to begin my Ph.D. at the University of Virginia
  • 05 Jun 2019: Received a scholarship to attend “Locomotion and Navigation from Flies to Robots” in Trieste, Italy (29 Apr – 10 May 2019; could not attend due to visa issues)
  • 09 Apr 2019: Two papers accepted at TENSYMP 2019
  • 03 Apr 2019: Received a research grant from the University Grants Commission (UGC), Bangladesh, for the project Domain Adaptive Transfer Learning for Multi-task Learning Models
  • 11 Mar 2019: Accepted into the Cornell, Maryland, Max Planck Pre-Doctoral School 2019 (05-10 Aug 2019), Saarbrücken, Germany
  • 04 Mar 2019: “A Gray Box Interpretable Visual Debugging Approach for Deep Sequence Learning Models” accepted at the ICLR-2019 Workshop on Debugging Machine Learning Models
  • 04 Mar 2019: Received a scholarship to attend the 7th Workshop on Collaborative Scientific Software Development at ICTP, Trieste, Italy (29 Apr – 10 May 2019)
  • 11 Feb 2019: “NEWSTRADCOIN: A Blockchain-Based Privacy-Preserving Secure News Trading Network” accepted at IC-BCT 2019
  • 22 Jan 2019: “Two-Phase Authentication and VPN-Based Secured Communication for IoT Home Networks” accepted at MIND-2019
  • 23 Nov 2018: “A Hierarchical Learning Model for Claim Validation” accepted at IJCCI-2018
  • 31 Aug 2018: “BARD: Bangla Article Classification using a New Comprehensive Dataset” accepted at ICBSLP-2018
  • 01 Jul 2018: Second Runner-up in the True Augmented Reality Challenge 2018
  • 15 Apr 2018: “iProtGly-SS: Identifying Protein Lysine Glycation Sites Using Sequence and Structure-Based Features” accepted at the Proteins Journal 2018
  • 22 Dec 2017: Received the NVIDIA Academic GPU Grant
  • 27 Nov 2017: “RAiTA: Recommending Accepted Answers using Textual Metadata” accepted at Springer IEMIS 2018
  • 18 Nov 2017: “VIM: A Big Data Analytics Tool for Data Visualization and Knowledge Mining” accepted at IEEE WIECON-ECE 2017
  • 11 Nov 2017: Attended the International Collegiate Programming Contest 2017 (Dhaka Regional) as a coach
  • 10 Sep 2017: Joined the Department of CSE, University of Dhaka
  • 09 Sep 2017: Conducted a workshop on IEEEXtreme 11.0 at United International University
  • 11 Jun 2017: Conducted a session on Kotlin at Google I/O Replay Dhaka 2017
  • 27 May 2017: Published “Mobile Cloud-Based Big Healthcare Data Processing in Smart Cities” in IEEE Access
  • 24-25 Mar 2017: Organized the Intra-University CSE Fest 2017 at United International University
  • 09-10 Mar 2017: Attended NCPC (National Collegiate Programming Contest) 2017 as a coach at CUET
  • 14 Jan 2017: Conducted a workshop on Android Mobile Application Development at the University of Dhaka
  • 7-9 Jan 2017: Attended the NSysS-2016 Conference at BUET, Dhaka, Bangladesh