At Booz Allen Hamilton, I built scalable RAG pipelines connecting enterprise knowledge bases with conversational AI, and tuned Pinecone, Milvus, and Weaviate retrieval layers. I also designed and deployed autonomous LLM agents for code refactoring and automated documentation.
I led R&D assessing LLM reasoning with Temporal Logics, and designed Azure experimentation, validation, and deployment workflows. Prompt optimization, caching, batching, and inference tuning reduced average response latency by 35% for production AI workflows.
At TTEC, I developed LSTM and Transformer models for customer sentiment across voice and text, and engineered chatbot flows to automate debt negotiations. I also deployed NLP models on AWS and contributed to integration with the customer service platform.
Earlier, at Slalom, I developed predictive models and REST APIs for enterprise finance and supply chain work. At AdventHealth, I automated public health data processing and built dashboards for disease monitoring and reporting.
