Design and deployment of Generative AI solutions using Azure OpenAI (GPT-4), LangChain, and RAG architectures to build intelligent support systems. Developed scalable RAG pipelines with Python, LlamaIndex, and vector databases to enable semantic search across enterprise knowledge bases. Processed large structured and unstructured datasets using Spark, Databricks, and Delta Lake to create production-ready ML assets. Deployed FastAPI services using Docker and Kubernetes while implementing MLflow-based monitoring, drift detection, and CI/CD pipelines to ensure stable and governed AI deployment.

At MiQ, I developed and deployed advanced machine learning solutions for digital advertising and audience analytics. I built Random Forest and XGBoost models for audience targeting and campaign optimization, increasing ad engagement across programmatic platforms. I designed scalable ML pipelines using Python and PySpark to process millions of campaign and behavioral records efficiently. Collaborating with cross-functional teams, I productionized ML models and automated deployment workflows, reducing release time by 30%. I implemented robust MLOps practices for model validation, monitoring, and drift detection, ensuring system reliability. Leveraging AWS services (EC2, S3), I enabled large-scale model training, deployment, and data processing to power campaign analytics and audience insights. Additionally, I applied techniques like classification, clustering, and predictive modeling to optimize omnichannel campaigns, contributing to a 15% ROI lift. I also designed and deployed CNN-based computer vision models using Python and TensorFlow to classify advertising creatives and extract visual features, improving targeting precision by 15%.

As a Python Developer at ZeOmega, I contributed to the Jiva healthcare platform by developing and maintaining backend components in Python, improving care management workflows by 25%. I designed and implemented RESTful APIs for secure healthcare data exchange using HL7/FHIR interoperability standards. My work included optimizing SQL Server queries, stored procedures, and data pipelines, boosting processing performance by 35% for large-scale healthcare datasets. Collaborating with cross-functional teams, I automated prior authorization and utilization management workflows using rule-based systems. I applied data analytics and basic ML techniques for patient risk stratification and outcome prediction. Additionally, I managed deployment and monitoring in both cloud and enterprise environments, ensuring 99.9% uptime, HIPAA compliance, and scalability.