Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am Bala Divya Udagandla, a production AI/ML Engineer focused on HIPAA-compliant healthcare LLM systems at enterprise payer scale. I have led production deployments at Optum/UnitedHealth Group and specialize in RAG pipelines, fine-tuning, and cost-efficient MLOps across AWS, Azure, and GCP. My work emphasizes robust, privacy-preserving AI workflows and measurable impact in healthcare outcomes. I excel in context engineering, LLM evaluation, and EHR integration with Epic and Cerner. I build scalable data pipelines with Docker, Kubernetes, Airflow, and distributed training tools, maintaining sub-150ms latency and 99.9% uptime while ensuring HIPAA/HITECH compliance and PHI de-identification.…I am Bala Divya Udagandla, a production AI/ML Engineer focused on HIPAA-compliant healthcare LLM systems at enterprise payer scale. I have led production deployments at Optum/UnitedHealth Group and specialize in RAG pipelines, fine-tuning, and cost-efficient MLOps across AWS, Azure, and GCP. My work emphasizes robust, privacy-preserving AI workflows and measurable impact in healthcare outcomes. I excel in context engineering, LLM evaluation, and EHR integration with Epic and Cerner. I build scalable data pipelines with Docker, Kubernetes, Airflow, and distributed training tools, maintaining sub-150ms latency and 99.9% uptime while ensuring HIPAA/HITECH compliance and PHI de-identification.

Bala Divya Udagandla

AI Engineer, Data Scientist, Full Stack Developer, +4





I am Bala Divya Udagandla, a production AI/ML Engineer focused on HIPAA-compliant healthcare LLM systems at enterprise payer scale. I have led production deployments at Optum/UnitedHealth Group and specialize in RAG pipelines, fine-tuning, and cost-efficient MLOps across AWS, Azure, and GCP. My work emphasizes robust, privacy-preserving AI workflows and measurable impact in healthcare outcomes. I excel in context engineering, LLM evaluation, and EHR integration with Epic and Cerner. I build scalable data pipelines with Docker, Kubernetes, Airflow, and distributed training tools, maintaining sub-150ms latency and 99.9% uptime while ensuring HIPAA/HITECH compliance and PHI de-identification.…I am Bala Divya Udagandla, a production AI/ML Engineer focused on HIPAA-compliant healthcare LLM systems at enterprise payer scale. I have led production deployments at Optum/UnitedHealth Group and specialize in RAG pipelines, fine-tuning, and cost-efficient MLOps across AWS, Azure, and GCP. My work emphasizes robust, privacy-preserving AI workflows and measurable impact in healthcare outcomes. I excel in context engineering, LLM evaluation, and EHR integration with Epic and Cerner. I build scalable data pipelines with Docker, Kubernetes, Airflow, and distributed training tools, maintaining sub-150ms latency and 99.9% uptime while ensuring HIPAA/HITECH compliance and PHI de-identification.

Available to hire

I am Bala Divya Udagandla, a production AI/ML Engineer focused on HIPAA-compliant healthcare LLM systems at enterprise payer scale. I have led production deployments at Optum/UnitedHealth Group and specialize in RAG pipelines, fine-tuning, and cost-efficient MLOps across AWS, Azure, and GCP. My work emphasizes robust, privacy-preserving AI workflows and measurable impact in healthcare outcomes.

I excel in context engineering, LLM evaluation, and EHR integration with Epic and Cerner. I build scalable data pipelines with Docker, Kubernetes, Airflow, and distributed training tools, maintaining sub-150ms latency and 99.9% uptime while ensuring HIPAA/HITECH compliance and PHI de-identification.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

AI/ML Engineer at Optum (UnitedHealth Group)

November 1, 2024 - Present

Designed and deployed classification and regression models for claims adjudication, member risk stratification, and 30-day readmission prediction using Python, PyTorch, and SQL on AWS SageMaker, achieving AUC > 0.90 and F1 > 0.85 on production datasets. Fine-tuned clinical LLMs (LoRA/QLoRA via Hugging Face) on proprietary EHR and claims corpora for named entity recognition and ICD-10 code suggestion; implemented A/B testing and output validation layers. Architected production RAG pipelines (LangChain, Pinecone, FAISS) for retrieval across millions of clinical notes and medical policies, reducing analyst review time by ~35% and optimizing token usage to reduce LLM costs by ~20%. Built multi-agent orchestration with LangGraph to automate care-gap identification, member record routing, and clinical alerting; ensured HIPAA/HITECH compliance and automated retraining at KPI degradation thresholds.

AI/ML Engineer at CitiusTech

April 1, 2021 - July 31, 2023

Delivered end-to-end ML solutions for U.S. payer and provider clients, including predictive models for patient no-show rates, chronic disease risk scoring, and claims denial prediction on EHR/FHIR data using Python and TensorFlow; improved clinical prediction accuracy by ~15%. Developed NLP pipelines with Hugging Face transformers and embeddings to extract diagnoses and clinical events from physician notes and discharge summaries with >90% entity-level accuracy. Implemented FHIR R4-compliant REST API integrations with Epic and Cerner, including Kafka-based streaming for continuous retraining on terabyte-scale clinical data; containerized services via Docker and Kubernetes on AWS/Azure ML with latency < 200 ms.

ML Engineer at Cybage Software

September 1, 2019 - March 31, 2021

Built supervised ML models (Random Forest, XGBoost, Logistic Regression) for demand forecasting, churn prediction, and sales propensity scoring across retail and BFSI clients, achieving KPI improvements of ~15%. Designed SQL-based ETL pipelines from Oracle DB and CRM systems into model-ready feature sets, reducing data prep effort by ~20%. Developed Power BI dashboards to track model outputs and KPIs, enabling data-driven decisions for delivery teams and clients.