Senior GenAI/ML Engineer with 12+ years of experience designing scalable Generative AI, Agentic AI, LLM, RAG, GraphRAG, and machine learning solutions across regulated industries. Hands-on with Azure OpenAI, Azure AI Agent Service, Azure AI Search, Claude/OpenAI, LangChain/LangGraph, and MCP to build enterprise-grade AI applications. I build production-ready, cloud-native AI services and pipelines (FastAPI, Docker, Kubernetes/AKS, vLLM, Redis), with strong LLMOps/MLOps for evaluation and observability (MLflow, LangSmith, DeepEval, RAGAS, OpenTelemetry). I also implement enterprise governance and security using Entra ID, Purview, and Key Vault to deliver compliant, secure AI systems for healthcare and financial domains.

Sai Ram PSr

Senior GenAI/ML Engineer with 12+ years of experience designing scalable Generative AI, Agentic AI, LLM, RAG, GraphRAG, and machine learning solutions across regulated industries. Hands-on with Azure OpenAI, Azure AI Agent Service, Azure AI Search, Claude/OpenAI, LangChain/LangGraph, and MCP to build enterprise-grade AI applications. I build production-ready, cloud-native AI services and pipelines (FastAPI, Docker, Kubernetes/AKS, vLLM, Redis), with strong LLMOps/MLOps for evaluation and observability (MLflow, LangSmith, DeepEval, RAGAS, OpenTelemetry). I also implement enterprise governance and security using Entra ID, Purview, and Key Vault to deliver compliant, secure AI systems for healthcare and financial domains.

Available to hire

Senior GenAI/ML Engineer with 12+ years of experience designing scalable Generative AI, Agentic AI, LLM, RAG, GraphRAG, and machine learning solutions across regulated industries. Hands-on with Azure OpenAI, Azure AI Agent Service, Azure AI Search, Claude/OpenAI, LangChain/LangGraph, and MCP to build enterprise-grade AI applications.

I build production-ready, cloud-native AI services and pipelines (FastAPI, Docker, Kubernetes/AKS, vLLM, Redis), with strong LLMOps/MLOps for evaluation and observability (MLflow, LangSmith, DeepEval, RAGAS, OpenTelemetry). I also implement enterprise governance and security using Entra ID, Purview, and Key Vault to deliver compliant, secure AI systems for healthcare and financial domains.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Beginner
See more

Work Experience

Senior GenAI/ML Engineer at Baylor Scott & White Health
February 1, 2025 - Present
Designed and delivered an enterprise Agentic AI & Clinical Knowledge Intelligence Platform. Built Azure Data Factory and PySpark pipelines to ingest EHR, FHIR, claims, laboratories, and clinical documents into AI-ready repositories. Implemented RAG with Azure AI Search (vector index) and Azure OpenAI embeddings, including Pinecone/Vector Index approaches for clinical knowledge retrieval. Developed document pipelines for classification, chunking, embedding, and indexing of clinical guidelines and patient records. Built LangChain workflows using tool/function calling with structured outputs and guardrails for search, summarization, and QA. Created LangGraph multi-agent workflows with agent memory for patient triage, utilization review, and care coordination. Orchestrated secure healthcare workflows using Azure AI Agent Service and integrated FHIR APIs through MCP. Implemented GraphRAG using Neo4j to connect diagnoses, medications, providers, and treatment plans. Added prompt engineering,
Senior GenAI/ML Engineer at Baylor Scott & White Health, Texas
February 1, 2025 - Present
Built enterprise agentic AI and clinical knowledge intelligence capabilities for healthcare operations and clinician decision support. Implemented Azure Data Factory and PySpark pipelines for EHR/FHIR/claims/documents ingestion into AI-ready repositories. Developed RAG and GraphRAG using Azure AI Search vector indexing, Azure OpenAI embeddings, Pinecone, and Neo4j for relationship discovery across clinical entities. Created LangChain/LangGraph workflows with tool/function calling, structured outputs, and agent memory for triage, utilization review, and care coordination. Orchestrated workflows using Azure AI Agent Service and integrated FHIR APIs through Model Context Protocol (MCP). Deployed scalable GenAI services via async FastAPI microservices, with evaluation and monitoring using MLflow/LangSmith/DeepEval/RAGAS and OpenTelemetry/Azure Monitor; secured workflows with Entra ID, Purview, Key Vault, and implemented guardrails plus caching to improve cost and quality.
GenAI Engineer at T Rowe Price
September 1, 2023 - January 31, 2025
Delivered enterprise GenAI solutions for investment research, portfolio analysis, and financial knowledge discovery. Built Azure Data Factory and PySpark pipelines for SEC filings, transcripts, analyst reports, and market data ingestion with incremental processing. Implemented RAG using Azure AI Search vector index, Azure OpenAI embeddings, FAISS, and tool-augmented LangChain workflows for summarization, question answering, and research retrieval. Used LangGraph multi-agent workflows with agent memory for automated research and analysis; integrated LiteLLM for model routing across Azure OpenAI and OpenAI. Reduced repeated LLM calls using prompt caching/semantic caching with Redis. Developed asynchronous FastAPI microservices and deployed on AKS with vLLM; orchestrated ingestion/inference using Airflow, Azure Functions, and Logic Apps with event-driven agents. Added observability with OpenTelemetry/Azure Monitor/MLflow and enforced security/compliance with Entra ID, Purview, and Key Vau
AI/ML Engineer at Kaiser Permanente
January 1, 2021 - August 31, 2023
Developed enterprise clinical AI and predictive analytics capabilities for readmission prediction, chronic disease management, and population health. Built AWS Glue and PySpark pipelines to ingest claims, EHR, pharmacy, laboratory, provider, and member datasets into S3; used AWS Glue/PySpark/Redshift for scalable transformations and dataset preparation. Performed EDA and feature engineering from diagnoses, medications, labs, encounter/utilization history. Trained LightGBM models for readmissions and disease progression/outcomes. Built K-Means clustering for member segmentation based on utilization/diagnosis/treatment patterns. Fine-tuned ClinicalBERT and BioBERT for extracting diagnoses/medications/procedures/clinical concepts from physician notes and discharge summaries. Applied SHAP and LIME for explainability. Evaluated models with Precision/Recall/F1/ROC-AUC/RMSE/MAE and selected models aligned with clinical and operational objectives. Implemented FastAPI inference services and dep
AI/ML Engineer at Kaiser Permanente, Maryland
January 1, 2021 - August 31, 2023
Developed enterprise clinical AI and predictive analytics for readmission prediction, chronic disease management, and population health. Built AWS Glue and PySpark pipelines to ingest claims, EHR, pharmacy, lab, provider, and member datasets into S3; transformed data using Spark/Redshift for predictive analytics. Engineered clinical features and trained LightGBM models for readmission and outcome prediction. Built K-Means clustering for member segmentation and fine-tuned ClinicalBERT/BioBERT for clinical concept extraction from notes and discharge summaries. Implemented model explainability (SHAP, LIME), evaluated with clinical metrics, and served models via FastAPI with AWS Lambda integration. Managed MLOps workflows using MLflow and Airflow with CI/CD; secured PHI using IAM/KMS/Lake Formation and monitored pipelines with CloudWatch.
Machine Learning Engineer at Optum
February 1, 2019 - December 31, 2020
Built healthcare predictive analytics and clinical risk modeling solutions. Developed PySpark pipelines using AWS Glue to consolidate claims, EHR, pharmacy, and provider data into S3. Engineered clinical and claims-based features from diagnosis codes, medications, utilization history, and labs to improve risk assessment. Built K-Means clustering for member segmentation and personalization. Implemented NLP extraction pipelines using spaCy and NLTK for diagnoses/procedures/medications/clinical observations from notes. Designed patient similarity approaches using clustering methods. Built preprocessing pipelines for missing values, class imbalance, categorical encoding, and scaling prior to training. Evaluated models using Precision/Recall/F1/ROC-AUC/RMSE/MAE and implemented SHAP explainability for clinician-facing insights. Created reusable modules for feature engineering, training, inference, and validation. Automated training/batch scoring/retraining using Airflow/MLflow/AWS CodePipeli
Machine Learning Engineer at Optum, Minnesota
February 1, 2019 - December 31, 2020
Created healthcare predictive analytics and clinical risk modeling solutions using claims/EHR/pharmacy/provider data. Built PySpark/Glue ingestion pipelines into S3 and performed preprocessing for imputation, encoding, scaling, and imbalance handling. Engineered clinical and claims-based features for risk assessment; developed clustering using K-Means for personalized care management. Implemented NLP pipelines using spaCy/NLTK for extracting medical entities from clinical text. Trained and evaluated multiple ML models and applied SHAP for clinician-facing explanations. Exposed models through FastAPI services and batch inference workflows integrated with Redshift reporting; automated training/scoring/retraining with Airflow/MLflow and AWS CI/CD. Ensured HIPAA compliance via IAM/KMS/encryption and monitored workloads via CloudWatch.
Python Developer at Credit One Bank
October 1, 2013 - October 1, 2018
Worked on enterprise credit card data integration and risk analytics. Developed Python scripts to ingest/process transaction, customer, merchant, and payment datasets from relational databases and flat files. Assisted in Python/SQL ETL workflows to cleanse, validate, and transform data for reporting and operational analysis. Wrote SQL queries, stored procedures, joins, views, and functions for customer/transaction/payment/portfolio reporting. Supported HDInsight Hadoop batch processing for large historical transaction/account volumes. Implemented incremental processing logic for newly created/updated records. Built Python validation/reconciliation scripts to compare source vs target datasets for accuracy and compliance needs (PCI-DSS/SOX). Supported dataset preparation for fraud monitoring/AML/operational risk analysis. Performed SQL tuning and performance improvements in Azure SQL Database and Synapse Analytics. Automated extraction and validation tasks using Python and shell scriptin
Python Developer at Credit One Bank, Hyderabad
October 1, 2013 - October 31, 2018
Built Python and SQL-based ETL/data integration for credit card transaction, customer, merchant, and payment datasets. Authored ETL workflows to cleanse/validate/transform data for reporting and operations. Wrote SQL queries, stored procedures, joins, views, and functions for customer/transaction/payment/portfolio reporting. Supported large-scale batch processing with HDInsight/Hadoop and implemented incremental data processing to capture new/updated records. Developed reconciliation/validation scripts to ensure data accuracy for PCI-DSS and SOX. Tuned SQL performance in Azure SQL Database and Synapse and automated extraction/file processing with Python/Shell. Monitored jobs with Azure Monitor and supported RBAC access controls for sensitive data; maintained reusable ETL utilities and logging/error handling components.

Education

Add your educational history here.

Qualifications

Industry Experience

Healthcare, Financial Services, Professional Services, Software & Internet

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Beginner
See more