Senior Software Engineer / Senior Machine Learning Engineer at ScienceSoft

Architected a multi-tenant retrieval-augmented generation (RAG) copilot on AWS with tenant-segregated ingestion of SharePoint and Google Drive data, embedding documents with transformer-based models and storing vectors in Pinecone (metadata in PostgreSQL) for secure Q&A. Built a hybrid dense/vector + keyword retrieval system; implemented LangChain-based query rewrites and late-chunking to optimize LLM response quality. Developed NLP pipelines for embeddings, semantic search, and context-aware query rewriting to improve retrieval accuracy. Containerized end-to-end solution (ingestion, Copilot UI, databases) and deployed on Amazon EKS behind Nginx with p95 latency <= 350 ms at 1M+ embeddings. Implemented ETL with Spark and Airflow; integrated SageMaker/Bedrock for model serving; automated deployment with Docker, Terraform, and GitHub Actions. Designed data schemas (PostgreSQL, Redshift, Snowflake) for analytics. Built FastAPI microservices for ingestion and OCR; monitored via CloudWatch,

Senior Software Engineer | Senior Machine Learning Engineer at ScienceSoft

Led the architecture of a multi-tenant RAG copilot on AWS integrating SharePoint and Google Drive data with transformer-based embeddings stored in Pinecone and PostgreSQL to support tenant-isolated Q&A. Developed hybrid retrieval systems using dense vector search combined with keyword indexing, and implemented LangChain-based optimizations enhancing LLM response quality. Containerized and deployed the solution on Amazon EKS ensuring low latency with large scale embedding operations. Built ETL pipelines with Apache Spark and Airflow and integrated scalable LLM model inference using Amazon SageMaker and Bedrock automated via Docker, Terraform, and CI/CD. Designed data storage schemas in Redshift and Snowflake for analytics, and introduced monitoring with CloudWatch, Prometheus, and Grafana. Mentored junior engineers in prompt engineering and MLOps best practices.

Senior Software Engineer / Senior Machine Learning Engineer at Tapcheck Inc.

Developed computer vision models for facial verification and liveness detection (PyTorch, OpenCV, MTCNN). Built fraud detection pipelines with XGBoost and LightGBM processing large transaction datasets in Spark/Hive. Designed feature engineering workflows in PySpark for time-series/behavioral data to improve fraud and churn prediction. Automated retraining/deployment with Airflow and Kubernetes, implementing CI/CD. Built hybrid edge-cloud inference using Docker and AWS Fargate for low-latency serving. Implemented monitoring dashboards in Grafana/Kibana for drift and performance; integrated ML APIs with backend/mobile teams; documented APIs and pipelines.

Senior Software Engineer | Senior Machine Learning Engineer at Tapcheck Inc.

Developed computer vision models focused on facial verification and liveness detection using PyTorch, OpenCV, and MTCNN. Designed and maintained fraud detection pipelines with XGBoost and LightGBM processing large transactional data with Spark and Hive. Implemented feature engineering workflows in PySpark to enhance predictive models for fraud and churn. Automated model retraining and deployments with Airflow and Kubernetes within CI/CD frameworks. Built hybrid edge-cloud inference systems on AWS Fargate for low latency serving. Developed monitoring systems in Grafana and Kibana for model and data drift detection and collaborated with cross-functional teams to integrate ML APIs and optimize real-time inference.

Software Engineer / Data Scientist / ML Engineer at Hallmark Health Care Solutions

Designed and implemented deep learning models for automated cell counting in pathology images using PyTorch/TensorFlow; applied optimization algorithms for medical device parameter tuning; implemented anomaly detection on sensor/camera data for hospital facility monitoring. Automated ETL pipelines for large-scale medical image and EHR datasets using Spark/Airflow; integrated conversational AI with EHR/CRM for scheduling and triage; implemented access controls for PHI compliance; built interactive dashboards with Tableau/Power BI; mentored juniors; provided production deployment guidance.