Enterprise applied research
2022 – PresentResearch Scientist · Dallas, Texas
- Led the design and deployment of a Retail Agentic Platform with Supervisor/ReAct orchestration, intent-based routing, specialized worker agents, a tool registry, Redis-backed session memory, and I/O guardrails on GKE.
- Delivered a full-stack Enterprise AI Platform on Kubernetes: React chat UI and BFF, vector knowledge API, document ingestion, LLM orchestration, and analytics. Consolidated nine microservice stubs into a single in-process orchestration layer.
- Architected a hybrid Azure and on-premises AI/ML platform that reduced data processing times by 90%.
- Proposed and prototyped computer vision models for automated checkout, smart carts, and shrink/loss prevention.
- Built and led a multidisciplinary AI team delivering a 50% improvement in threat detection and multi-million-dollar cost avoidance.
- Shaped the enterprise AI R&D roadmap and aligned initiatives with C-suite business priorities.