Sri Venkatesh Kumar S.

About Sri Venkatesh Kumar S.

I’m a Senior Generative AI, Machine Learning, and Data Engineering professional with over 8 years of experience delivering AI-powered solutions across healthcare, legal tech, telecom, and enterprise systems. I specialize in building and deploying LLM-based, multi-agent architectures using tools like LangGraph, LangChain, and RAG pipelines integrated with T5, BERT, and GPT-based models.I’ve designed autonomous AI agents that perform retrieval, summarization, reasoning, and validation, all orchestrated through graph-based workflows and deployed on cloud-native infrastructure (Docker, Kubernetes, GCP Vertex AI, Cloud Run). These solutions are supported by strong CI/CD pipelines, real-time monitoring, and explainability layers using SHAP and model drift detection.Beyond GenAI, I bring deep expertise in:Data Engineering: Building real-time & batch data pipelines using Apache Spark, Kafka, BigQuery, Hive, and SQL/NoSQL platforms.Data Science & ML: Developing predictive models using XGBoost, Prophet, Scikit-Learn, TensorFlow, and PyTorch, with experience in feature engineering, hyperparameter tuning, and A/B testing.MLOps & DevOps: End-to-end deployment pipelines with GitHub Actions, Dockerized ML APIs, FastAPI, and automated model retraining & rollback strategies on cloud.Whether it\'s fine-tuning ClinicalBERT for EHR summarization, enhancing CT/MRI image quality with diffusion models, or developing AI copilots with LangGraph, I bridge research and production to deliver scalable, ethical, and impactful AI solutions.

Experience

1. Data Scientist, Gen Ai Ml

   CVS Health

   Feb 2024 — Present · RI, US

   Developed and trained Generative AI models (GANs, Diffusion Models) for medical imaging, enhancing diagnostic accuracy.Developed a LangGraph-powered GenAI agent framework integrated with BioBERT and ClinicalT5 for intelligent querying of medical knowledge graphs.Connected LLMs to Neo4j graph database for real-time reasoning over disease–gene–drug relationships, adverse drug reaction predictions, and biomarker similarity searches.Deployed APIs with FastAPI, Docker, and GCP Cloud Run, enabling integration into CVS telehealth systems for patient risk scoring and decision support.Built monitoring and explainability tools using SHAP, ensuring transparency and compliance with HIPAA standards.Implemented transformer models (BERT, T5) to summarize electronic health records, reducing clinical documentation time.Enhanced telemedicine services with LLM-powered agents (LangChain), improving patient interaction and health information access.Fine-tuned NLP models (LaMDA, BLOOM) for AI-driven communication and content generation.Deployed scalable AI/ML models using AWS Bedrock, Docker, and Kubernetes, ensuring high reliability and availability.Monitored AI models in real-time using the ELK Stack to detect data drift and maintain accuracy in diagnosis.Built health data knowledge graphs with embeddings and Pinecone, improving medical research retrieval.Designed scalable solutions for genomic analysis using Hadoop and PySpark, increasing throughput for large datasets.Developed NLP chatbots with spaCy for symptom-checker features, improving patient satisfaction.Fine-tuned transformer models for semantic search in clinical trial data, Created synthetic health data pipelines for testing and research while ensuring patient privacy.Improved model interpretability using SHAP and LIME, enabling explainable AI for clinicians.Configured automated ETL pipelines with Apache NiFi and Airflow.Developed AI systems for real-time anomaly detection in patient health metrics.

Education

• University of South Florida

  Master's degree, Business Analytics and information systems

• Lakireddy Bali Reddy College of Engineering(Autonomous)

  Bachelor of Technology - BTech