Nagendar Goud
I’m a senior AI/ML engineer with 11+ years of experience building production-grade machine learning and generative AI solutions for insurance, healthcare, and financial services.
About
I’m a senior AI/ML engineer with 11+ years of experience building production-grade machine learning and generative AI solutions for insurance, healthcare, and financial services. I specialize in designing scalable ML pipelines, developing interpretable predictive models, and delivering end-to-end ML services—from data ingestion and feature engineering to training, evaluation, deployment, and monitoring.
I also build responsible, human-in-the-loop AI systems, including retrieval-augmented generation (RAG) and agentic workflows that integrate LLMs with enterprise tools and secure, governed data access. Whether it’s underwriting risk, fraud and waste analytics, or advisor/investment decision intelligence, I focus on measurable business outcomes, strong data quality practices, and reliable model operations in cloud environments like AWS, Azure, and GCP.
- 11 years' experience
- Joined October 2026
Industry experience
- Financial Services
- Healthcare
- Professional Services
Work experience· 5 roles
Engineered an insurance risk prediction and underwriting analytics platform using Python, PySpark, XGBoost, and AWS SageMaker. Reduced predicted default probability by 12% and underwriting review time by 30% by translating underwriting requirements into ML workflows and evidence-driven review experiences. Built a production RAG solution (LangChain, Qdrant/Elasticsearch, LangSmith/Ragas) to retrieve underwriting guidelines and generate source-grounded summaries for reviewers. Developed agentic LangGraph-based workflows with tool calling and human-in-the-loop checkpoints, integrated via FastAPI microservices and SageMaker model-serving endpoints. Implemented data lineage documentation, reusable feature engineering/validation pipelines, and ML governance through versioned artifacts, MLflow, and CloudWatch monitoring; containerized and deployed services with Docker/ECR/EKS; automated CI/CD and infrastructure provisioning with GitHub Actions and Terraform.
Developed pharmacy claims Fraud, Waste & Abuse (FWA) risk analytics using Python, SQL, PySpark, scikit-learn, XGBoost, and statistical methods, improving investigation prioritization by 15%. Built end-to-end analytics workflows connecting pharmacy claims and reference data across member, provider, prescriber, and drug dimensions. Prepared large-scale datasets by resolving duplicates/reversals/missing values and temporal anomalies, then engineered claim/utilization/provider/drug features for suspicious billing behavior detection. Trained and tuned supervised models (XGBoost, Random Forest, Logistic Regression) and evaluated neural approaches; used SHAP for explainability and threshold optimization for operational precision/recall balance. Implemented RAG with LangChain and vector retrieval for policy/guideline and historical case evidence, and integrated Azure OpenAI for source-grounded investigation summaries. Established monitoring for data/model drift, built agent workflows for evide
Engineered an advisor trading analytics and investment decision intelligence platform using Python, PySpark, SQL, scikit-learn, and XGBoost, reducing advisor analytics turnaround time by 25%. Built SQL/PySpark ETL pipelines to join trading, execution, account, position, security, and performance data with validation and reconciliation controls. Reconciled and standardized identifiers, constructed portfolio state and trading-event sequences, and engineered advisor-level behavioral features (frequency, turnover, concentration, allocation changes). Applied clustering/statistical analysis to segment advisor behavior profiles and developed supervised classifiers to identify trading behaviors relevant for portfolio analytics. Implemented batch inference with PySpark and exposed results via FastAPI REST APIs deployed with Docker on Google Compute Engine. Supported Vertex AI training/evaluation workflows with repeatable artifacts and monitoring through Google Cloud; improved release quality wi
Education
Hire Nagendar Goud today
Post a job and invite Nagendar, or contact them directly. No hidden commission.