Who Am I?

Hello! I'm Adib, a Machine Learning researcher at Vector Institute, University Health Network, INSERM, and BorealisAI, as well as a Computer Science and Neuroscience student at the University of Toronto. My passion lies in the realm of Intelligence, encompassing both the fascinating intricacies of the Human Brain and the evolving landscape of Artificial Intelligence. I am ambitious to push the boundaries of developing Artificial Intelligence and understanding Biological Intelligence.

My expertise is in developing cutting-edge machine learning models in healthcare, neuroscience, and biology. I've first-authored a paper on EHRMamba, a state-of-the-art foundation model for electronic health records, under review at NeurIPS ML4H 2024, accompanied by Odyssey, the largest open-source toolkit for full-stack EHR modeling. I've also engineered CodonTransformer, a widely-used model that optimizes DNA sequences for protein expression in 164 organisms, with 2,000+ downloads and 88+ GitHub stars, under review at Nature Communications.

Apart from my AI enthusiasm, I love building my community. I've founded the Google Developer Student Club at my campus, along with Canada's largest AI hackathon, GenAI Genesis, attracting 1,200+ applicants from 10+ countries.

What i'm doing

  • Machine Learning

    Machine Learning

    Leveraging cutting-edge NLP, Computer Vision,
    & ML techniques to pioneer innovative solutions.

  • Neuroscience

    AI In Healthcare

    Developing state-of-the-art AI models for healthcare applications at Vector Institute and UHN.

  • Google

    Community Development

    Leading the Google Developer Student Club, organizing Canada's largest AI hackathon.

  • Archery

    Archery

    Finding solace, growth, and fulfillment through
    the art of shooting arrows with a bow.

Resume

Download Resume

EDUCATION

  1. University of Toronto

    Sep 2022 - Expected June 2026

    Honours B.Sc. in (Co-op) Computer Science and Neuroscience; cGPA: 3.96/4.0 | TA: Python, Probability, Discrete Mathematics

EXPERIENCE

  1. Applied Machine Learning Engineer | Vector Institute

    Jan 2024 - Present

    • First-authored a research paper at NeurIPS ML4H 2024, vectorinstitute.github.io/EHRMamba, under review, introducing EHRMamba, a state-of-the-art foundation model for scalable modeling of Electronic Health Records.

    • Developed EHRMamba, the first model to excel in both EHR forecasting and 6 clinical prediction tasks, using a scalable architecture that processes 300% longer patient histories with 38% higher compute efficiency, accelerating AI integration in healthcare systems.

    • Engineered Odyssey, a pioneering open-source toolkit for full-stack EHR modeling, encompassing novel medical data tokenization, PyTorch-based Transformer and Mamba architectures, and robust tools for EHR forecasting, clinical prediction, and interpretability.

  2. Machine Learning Researcher | INSERM, France

    July 2023 – September 2024

    • Engineered CodonTransformer, a state-of-the-art model with 2,000+ downloads that optimizes DNA sequences for protein expression in 164 organisms, trained on a dataset of 1,000,000+ gene-protein pairs with PyTorch Lightning.

    • First-authored a manuscript at Nature Communications, adibvafa.github.io/CodonTransformer, under review, showcasing estimation to save $5,000,000+ with CodonTransformer in vaccine manufacturing and protein production.

    • Built the largest open-source library with 88+ GitHub stars for codon optimization, providing a full-stack ML pipeline.

  3. Computer Vision Researcher | BorealisAI

    March 2024 - May 2024

    • Developed a pioneering image-to-text multimodal AI system, github.com/Adibvafa/RadioCareBorealisAI, for chest X-ray radiology report generation, tackling critical healthcare accessibility gaps for 3.8 million rural Canadians.

    • Architected a cutting-edge medical imaging model by integrating Vision Transformer, BLIP, and GPT technologies, leveraging over 377,000 chest x-ray images to surpass previous benchmarks with an 86.27% precision and 83.28% F1-score.

PROJECTS

  1. UnsearchableTO | New Builds Hackathon | GitHub Website Twitter

    Python, Next.js, FastAPI, Gemini

    • Made the entire city of Toronto searchable with over half a million objects using Python, Next.js, FastAPI, Gemini, and Cohere API.
    • Reached 32,000+ views on X, 2,000+ user queries, and 14,000+ searched locations in 2 days after launch.

  2. Vector Institute's Cyclops: Stroke Prediction | GitHub Kaggle

    TensorFlow, Scikit-Learn, Pandas

    • Enhanced Cyclops, a 100,000+ line ML toolkit for healthcare research, by resolving critical deployment issues.
    • Authored a Kaggle tutorial (3,000+ views) on Cyclops for stroke prediction with 90% accuracy using TensorFlow, Sklearn, & Pandas.

  3. Hugging Face Contributor | GitHub

    NLP, Open-source Development

    • Integrated attention-hybrid Mamba2 models into Hugging Face's Transformers library, the industry-standard open-source library for Natural Language Processing, expanding the library's cutting-edge offerings for millions of AI practitioners worldwide.

LEADERSHIP

  1. Founder & Lead | Google Developer Student Club (GDSC)

    Aug 2023 - Present

    • Founded GenAI Genesis, Canada's largest AI hackathon, with 1,200+ applicants from 10+ countries, sponsored by Google.

CERTIFICATES

  1. Machine Learning & Deep Learning Specializations | Coursera, DeepLearning.ai, Stanford Online

    2023

    • Aced 8 courses in Natural Language Processing, Computer Vision, and Reinforcement Learning using TensorFlow and Keras.

  2. International Biology Olympiad – Silver Medal | 32nd IBO, Lisbon, Portugal

    2021

    • Achieved Silver Medal in the world's most prominent biology competition among 300 students from 72 countries.

SKILLS

  1. Languages

    Python, Java, C/C++, HTML, CSS, JavaScript, TypeScript, SQL, Assembly, Shell, Bash

  2. Frameworks

    PyTorch, TensorFlow, Keras, Hugging Face, JAX, FastAPI, React, Next.js, OpenCV, XLA, NumPy, Pandas, Scikit-Learn

  3. Tools & Cloud

    GCP, AWS, Azure, CUDA, SLURM, Git, MySQL, Firebase, Supabase, MongoDB, Unix, Linux, PowerBI, Vercel, Jupyter

Projects

Contact

I am eager to connect and talk!