Vaibhav Bhardwaj

I’m an AI/ML Engineer with 2 years of experience designing and deploying end-to-end machine learning systems aligned with business objectives. I focus on the full ML lifecycle—from data collection and preprocessing to feature engineering, model training and fine-tuning (PEFT/LoRA), evaluation, and REST API deployment.

At NetEdge, I help deliver real-world outcomes by building an XAI-enabled deep learning pipeline (UNet + CNN) for neonatal brain MRI segmentation and neurodevelopmental outcome prediction. My work reduced radiologist review time by ~40%, translating model performance into measurable clinical value.

I also built an AI-assisted Clinical Recall System using an end-to-end RAG pipeline with semantic chunking, FAISS vector search, and sentence-transformer embeddings. By deploying a Dockerized FastAPI backend and validating with load testing, I reduced retrieval time from ~8 minutes to <10 seconds (98% reduction) while achieving >90% retrieval relevance across 500+ clinical notes.

Earlier, as a Machine Learning Intern and freelance Data Analyst, I strengthened my statistical and data-quality instincts—using methods like hypothesis testing, PCA, and SMOTE to keep models aligned with objectives. I carry that mindset into my projects too, from RADBert (BERT + PEFT/LoRA for radiology report entity extraction) to TransGuard (imbalanced fraud detection with an MLOps stack including MLflow and Airflow), always aiming for reliable, deployable impact.