Architected and shipped StatVisor, a production-grade multi-agent AI system using LangGraph, LangChain, LlamaIndex, Azure OpenAI GPT-4, and FAISS-backed RAG pipelines — engineered prompt engineering, function/tool calling, context grounding, BM25 sparse + dense hybrid retrieval, and OCR-based document ingestion pipelines for context-aware extraction across large enterprise datasets

• Built end-to-end LLM governance and observability infrastructure using FastAPI, Azure Monitor, and MCP — enforced audit logging, cost controls, schema validation, and responsible AI compliance across production AI systems

• Built and evaluated advanced ML models using XGBoost and scikit-learn — produced performance dashboards using Matplotlib and Seaborn for senior stakeholder reporting and data-driven decision making

• Engineered cloud data pipelines on Snowflake — queried and transformed data using Snowflake Cortex LLM functions, vector search, and Cortex Analyst for AI-native retrieval workflows; built and maintained cross-platform pipelines across BigQuery and Redshift

• Orchestrated multi-step agentic workflows using LangGraph, LangChain, LlamaIndex, and Semantic Kernel with guardrails, multi-hop reasoning, and tool-calling agents; implemented data governance and metadata management frameworks aligned with enterprise cataloging standards (Collibra); built graph-based retrieval pipelines using Neo4j and Stardog for relationship-aware knowledge graph use cases

• Built multiple production R tools and statistical REST APIs using R, Plumber, ggplot2, dplyr, and tidyverse; engineered async Python (FastAPI) backend services with complex SQL across Snowflake, BigQuery, and Redshift; presented AI and statistical solutions directly to senior P&G leadership, translating complex system outputs into governance-aligned business decisions

• Built and deployed AI/ML models and generative AI agents in Python and R automating supply chain insights and actions — implemented hallucination mitigation strategies including output validation, grounding, and confidence scoring; developed production-grade pipelines with feature engineering and model versioning across planning, logistics, and inventory optimization workflows

• Architected scalable data and analytics solutions on Google Cloud using Vertex AI — trained and deployed ML models into enterprise supply chain platforms; designed high-throughput Python data pipelines with concurrency and load-balancing algorithms reducing processing overhead by 35% across millions of supply chain records

• Built automated multi-threaded scraping (Selenium) and scoring system processing 11K+ sources in under 3 seconds with 95%+ accuracy — applied heuristics and optimization techniques for warehouse resource utilization, lane allocation, and AI-driven scheduling; adopted as a framework standard by 15+ engineers across 2 teams

• Developed Power BI dashboards with DAX and semantic layer modeling for supply chain KPI tracking; orchestrated end-to-end ML workflows using Vertex AI, Airflow, MLflow, Docker, Kubernetes, Terraform, Spark, and Kafka for scalable, production-ready data infrastructure

• Built and optimized large-scale data pipelines using PySpark, Delta Lake, and Python — performed data analysis, validation, and SQL operations on Delta tables; tracked experiments and managed model lifecycle using MLflow ensuring quality and reliability across high-volume datasets

• Engineered and orchestrated end-to-end data workflows on Azure Cloud using ADF and Airflow — leveraged Azure compute and storage services with Docker-containerized services for scalable, production-grade data engineering and MLOps workloads

• Collaborated with cross-functional teams to design and deliver data solutions — implemented CI/CD practices for pipeline deployments, monitored workflow observability, and optimized pipeline performance across the full data lifecycle using Azure DevOps and Git