[]
Machine Learning Research Engineer | MSc ML & Data Mining | Applying AI in Healthcare & Sustainability
- 🎓 MSc in Machine Learning & Data Mining at Université Jean Monnet, France
- 🔬 Research Engineer at Laboratoire Coactis - working on NLP for sustainability discourse and Healthcare AI
- 📊 Built explainable ML systems for ICU infection risk prediction (AUC-ROC > 0.98)
- 🤖 Specialized in NLP, LLMs, Multi-Agent Systems, and Explainable AI
- 📄 Paper under review: Explainable ICU Infection Risk Prediction (Int. J. Medical Informatics)
- Multi-agent LLM systems with LangGraph
- Physics-informed machine learning
- Explainable AI for high-stakes domains
- Graph Neural Networks for recommendations
Languages: Python, C++, SQL, R
ML/DL: PyTorch, TensorFlow, Scikit-learn, HuggingFace, LangChain, LangGraph
Specializations: NLP/LLMs (BERT, GPT, RAG, LoRA), Computer Vision, GNNs, Transfer Learning, Multi-Agent Systems
MLOps: Docker, Kubernetes, SLURM, MLflow, Optuna, W&B, DVC, Git
Data: Pandas, NumPy, SHAP, LIME, PostgreSQL, MongoDB, Vector DBs (Pinecone, Chroma)
Cloud/Viz: Azure, CUDA, Streamlit, Plotly, Matplotlib
- French Sustainability Paradox Detection: Built first large-scale corpus (1,074 videos) with 97% accuracy using NLP + LLMs
- ICU Infection Risk Prediction: Explainable ML pipeline integrating clinical and organizational data
- Graph Neural Networks for E-Commerce: Kaggle competition - benchmarked NGCF, LightGCN (Hitrate@10: 0.36)
- Multilingual LLM Bias Evaluation: Quantified translation bias across 11 languages using XQuAD
- CryoET 3D Object Detection: Protein complex detection in cryo-electron tomograms
- LinkedIn: zahirahmadchaudhry
- Email: zahirahmadchaudhry[at]gmail[dot]com
- Kaggle: zahirahmadch