Bala Rama Krishna Generative AI Engineer

GenAI Engineer

Vatica Health

Aug 2024 - Current (1 year 11 months)

• Designed and built a HIPAA-compliant clinical RAG platform on Microsoft Azure to support HCC medical risk coding and clinical documentation improvement workflows. • Designed the end-to-end RAG pipeline architecture including document ingestion, preprocessing, chunking, embedding generation, vector indexing, query routing, retrieval, context assembly, and LLM-grounded response generation. • Built a document ingestion and preprocessing pipeline for clinical notes, discharge summaries, and EHR exports, applying PHI masking via Microsoft Presidio. • Implemented hybrid retrieval combining dense vector search and BM25 keyword retrieval over Weaviate and Azure AI Search. • Integrated re-ranking using cross-encoder models and applied context comp

AI/ML Engineer

Wells Fargo

Jun 2022 - Jul 2024 (2 years 2 months)

• Designed and delivered a large-scale Azure-first machine learning platform to automate credit risk decisioning and detect financial fraud in real time. • Built and orchestrated distributed model training and inference pipelines using PyTorch, PySpark MLlib, and Azure Databricks. • Built credit risk scoring models using XGBoost, LightGBM, and ensemble techniques achieving accuracy above 89%. • Performed fairness testing using Fairlearn and Aequitas to ensure AI-driven fair lending compliance. • Implemented SHAP analysis for model explainability to meet CFPB requirements. • Automated model deployment via GitHub Actions CI/CD pipelines with performance gates and automatic rollback.

ML Engineer / Data Scientist

Target Corporation

Sep 2020 - May 2022 (1 year 9 months)

• Built an end-to-end AI-powered Customer Personalization and Demand Intelligence platform on AWS. • Designed a collaborative filtering recommendation engine on AWS SageMaker to surface personalized product suggestions. • Developed customer segmentation models using K-Means clustering. • Built time-series demand forecasting models using ARIMA and Prophet for workforce planning. • Developed an anomaly detection pipeline using Isolation Forest for fraud analytics. • Built automated data pipelines using AWS Glue, PySpark, and SQL.

Data Scientist

Liberty Mutual Insurance

Mar 2019 - Aug 2020 (1 year 6 months)

• Built end-to-end AI pipelines on AWS for insurance claims processing, underwriting risk scoring, and intelligent document extraction. • Designed intelligent document processing systems using NLP (spaCy, NLTK) to automate extraction from unstructured policy and claim documents. • Built computer vision pipelines using CNNs and OpenCV for property damage assessment. • Developed underwriting risk scoring models using ensemble techniques like Random Forest and XGBoost.

Software Engineer — Data Science & NLP

Hexaware Technologies

Jun 2016 - Nov 2018 (2 years 6 months)

• Developed an AI-driven Customer Sentiment Analytics and Churn Prediction platform on Microsoft Azure. • Built sentiment analysis models using NLP techniques (Word2Vec, GloVe) with over 85% accuracy. • Applied topic modeling (LDA) and named entity recognition to identify recurring themes in customer feedback. • Built customer segmentation models using K-Means and hierarchical clustering. • Developed REST API services for real-time sentiment scoring and churn prediction.

Bala Rama Krishna

About

Detail

Timeline

Résumé