Senior Generative AI Data Engineer at Charles Schwab Corporation

Designed and led architecture for an enterprise RAG and agentic AI support platform using Azure OpenAI, Azure AI Search, LangChain, and LangGraph. Built scalable ingestion and document processing pipelines using Azure Data Factory, Azure Functions, Azure Databricks, and PySpark, including validation, enrichment, deduplication, and semantic chunking. Implemented hybrid search (vector + keyword + metadata filtering + reranking) and asynchronous FastAPI microservices to expose retrieval, embedding, search, summarization, and agent capabilities. Engineered multi-step agent workflows for classification, policy validation, summarization, tool execution, and controlled human review. Developed reusable MCP-style connectors for standardized tool calling and integrations. Established data quality and evaluation frameworks for ingestion outputs and retrieval relevance; containerized services with Docker and deployed on AKS with CI/CD and monitoring via Azure DevOps/GitHub Actions and Azure Monito

Senior AI /ML Data Engineer at Johnson & Johnson

Built a clinical data insights and document intelligence platform integrating clinical, claims, and provider datasets for analytics and ML. Developed clinical risk scoring and patient segmentation models using XGBoost and scikit-learn with lifecycle tracking in MLflow and deployment via Amazon SageMaker. Implemented clinical NLP and document intelligence workflows using spaCy, ClinicalBERT, BioBERT, Amazon Textract, and Amazon Comprehend Medical for entity extraction and structured medical data from unstructured sources. Introduced a GenAI retrieval-augmented clinical search and summarization layer using Amazon Bedrock, LangChain, and OpenSearch. Delivered production services and orchestration on AWS using Lambda, Step Functions, API Gateway, CloudWatch; used Docker and CI/CD for reliability, scalability, and observability.

AI/ML Data Engineer at JPMorgan Chase

Developed reusable Dataflow, Pub/Sub, BigQuery, and Python pipeline components for customer and portfolio analytics. Built segmentation, churn prediction, propensity scoring, and portfolio trend analysis models using XGBoost, TensorFlow, and Vertex AI. Created standardized feature engineering datasets in BigQuery to support repeatable analytics, training, batch scoring, and reporting. Added financial text analytics using BERT/FinBERT for sentiment/topic/intent extraction. Built backend services with FastAPI/Cloud Functions to expose model scores and metrics to internal applications; automated recurring refresh and scoring workflows using Cloud Composer/Scheduler/Functions. Supported production releases with Cloud Build, Artifact Registry, logging/monitoring, and Git-based CI/CD.

ML Engineer at Publix

Built Python text-processing pipelines for customer reviews, surveys, product comments, and store-level feedback. Implemented ingestion using Azure Data Factory, Azure Blob Storage, and Azure SQL Database; processed larger datasets with Azure Databricks and PySpark for cleansing, joins, deduplication, and analytical table preparation. Developed sentiment and topic modeling workflows using spaCy, NLTK, scikit-learn, and BERT; applied TF-IDF, LDA, clustering, and rule-based features for issue/theme identification. Used Azure Cognitive Services for entity/key phrase extraction where applicable. Built batch training/scoring workflows with Azure ML and MLflow, and created Flask-based REST APIs for distributing results to dashboards. Improved turnaround time and accuracy while reducing manual review effort.