hey, |
I mess around with AI and biology sometimes.
I'm researching how we can use transformers and multi-modal learning to catch neurodegenerative diseases earlier. Also into bioinformatics, NLP, and making AI systems explainable.
I'm researching how we can use transformers and multi-modal learning to catch neurodegenerative diseases earlier. Also into bioinformatics, NLP, and making AI systems explainable.
I'm Fahad Nadim Ziad, a final-year Computer Science student at BRAC University with a deep focus on building intelligent systems for scientific domains. My journey into AI began with a fascination for unraveling complex systems, which has evolved into a dedicated pursuit of applying machine learning to high-stakes problems in biology and healthcare.
Across my research, I've developed a strong focus on the importance of model transparency. I believe that for AI to be truly useful in science, we must be able to trust its reasoning. This has led me to consistently incorporate explainable AI (XAI) techniques like SHAP and LIME into my work, ensuring my models provide not just answers, but also interpretable insights.
My goal is to build AI that is not just a tool, but a trustworthy partner in discovery. I operate on the principle that the best solutions are not only accurate but also robust and interpretable. My methodology involves breaking down complex challenges—like handling incomplete clinical data or aligning a language model with biological knowledge—and architecting targeted, explainable systems from the ground up.
BRAC University
Dhaka, Bangladesh • May 2022 - February 2026
| CSE443: Bioinformatics I | CSE440: Natural Language Processing II |
| CSE427: Machine Learning | CSE422: Artificial Intelligence |
| CSE424: Pattern Recognition | CSE428: Image Processing |
| CSE402: Optimization | CSE331: Automata and Computability |
| CSE321: Operating Systems | CSE420: Compiler Design |
| MATH216: Linear Algebra & Fourier Analysis | STA201: Elements of Statistics and Probability |
Birshreshtha Noor Mohammad Public College
Dhaka • 2021
Willes Little Flower School & College
Dhaka • 2019
This is a selection of my hypothesis-driven research work, where the primary goal was to investigate a concept, develop a new methodology, or provide a detailed proof-of-concept.
B.Sc. Thesis (Phase 3 in Progress)
Architecting a novel, modality-aware Transformer framework to diagnose neurodegenerative diseases from multi-modal neuroimaging data. This work pioneers a 'modality dropout' scheme to build a system robust to the real-world challenge of incomplete clinical data.
[read full case study]
CSE443: Bioinformatics
Led a team research project developing a novel Knowledge Graph Separation methodology for quantifying structural alignment of LLM embeddings with biomedical knowledge graphs following QLoRA fine-tuning. Demonstrated 126% improvement in embedding geometric alignment with Phi-3 Mini.
BRAC University Business Club
Executed multi-platform digital marketing campaigns to promote club events, significantly increasing student attendance and online visibility.
IABC | BRACU
Organized Model UN events, led club promotion initiatives, and managed member recruitment drives.
Robotics Club of BRAC University
Participated in foundational engineering workshops and contributed to project planning for competitive technical festivals.
BRAC University Film Club
Hosted and emceed club events, including film screenings and workshops, facilitating audience engagement.
Willes Science Club
Led planning and execution of large-scale science and IT events, mentoring junior members and coordinating educational workshops.
This is a gallery of my technical and software engineering projects, where the focus was on building functional systems, implementing complex algorithms, or demonstrating specific technical skills.
CSE420: Compiler Design (Fall 2025)
Architecting and implementing a functional compiler for a subset of the C programming language. The project covers the full compiler pipeline, from lexical analysis to final code generation.
CSE321: Operating Systems
Single-handedly designed and implemented a complete virtual file system from scratch in C, featuring custom superblock management with CRC32 checksums and a Unix-style inode system.
CSE424: Pattern Recognition
Developed a stacking ensemble model that achieved 96% accuracy and 100% recall in forecasting U.S. recessions. Used SHAP to interpret the model and validate its economic logic.
I'm always open to discussing research opportunities, collaborations, or interesting projects. Feel free to reach out through any of the channels below.
Location
Dhaka, Bangladesh
💡 Currently available for research collaborations, AI/ML projects, and software engineering opportunities.