GenAI Engineer at Broadridge Financial Solutions

Architected and deployed enterprise-scale Generative AI solutions leveraging GPT and open-source LLMs to improve scientific document analysis and research workflow efficiency by 40%+. Designed and implemented RAG pipelines with Pinecone, FAISS, and Weaviate to boost response accuracy and reduce hallucinations by 30%+. Led fine-tuning of foundation models on clinical trial and biomedical datasets, improving domain relevance, compliance, and model performance in regulated environments. Built scalable AI microservices with Python, FastAPI, Docker, and Kubernetes enabling real-time inference with latency under 200ms. Developed prompt engineering frameworks and automated evaluation pipelines. Implemented end-to-end MLOps pipelines (MLflow, CI/CD) for model training, versioning, deployment, monitoring, and rollback. Integrated GenAI apps with data lakes, APIs, and knowledge graphs for advanced analytics and real-time decision support. Optimized inference via quantization and batching, reduci

AI/ML Engineer at IBM

Designed and deployed scalable ML models for trading analytics, risk prediction, and customer segmentation, improving decision accuracy by 20%+. Led end-to-end ML lifecycle (data ingestion → modeling → deployment). Built and optimized Spark-based data pipelines on AWS/Azure, reducing model training time by 30%+. Developed deep learning and NLP models (TensorFlow, PyTorch) for anomaly detection and predictive analytics, increasing forecasting accuracy by 15-25%. Deployed ML models as REST APIs using FastAPI/Flask with Docker and Kubernetes for real-time inference. Automated feature engineering with Airflow, reducing manual effort by 40%. Implemented model monitoring and drift detection to ensure reliability and governance.