Alex
Bates

Deployed enterprise AI full-stack applications with Python, FastAPI, React, Next.js, and TypeScript across Azure and AWS, reducing client onboarding time by 20% while supporting scalable production workflows. Delivered production generative AI applications on Azure OpenAI, OpenAI, and Amazon Bedrock — intelligent document search and conversational experiences that improved user query satisfaction by 30%. Architected end-to-end RAG pipelines for enterprise document repositories spanning ingestion, metadata extraction, chunking, embedding generation, retrieval, ranking, and LLM response generation, improving retrieval accuracy from 62% to 87%. Accelerated Azure AI Search with vector search, keyword search, semantic ranking, metadata filtering, and relevance tuning to cut average search latency by 20% across large document collections. Engineered automated AI workflows with LangChain, LangGraph, Python, FastAPI, and Azure Functions, integrating LLMs, enterprise APIs, and decision-making tools to streamline the deployment pipeline and cut manual orchestration effort for the team. Led design discussions, code reviews, and production troubleshooting with Docker, Kubernetes, Terraform, Databricks, MLflow, Jenkins CI/CD, and Prometheus, delivering faster deployment cycles and higher service reliability.

Implemented enterprise full-stack applications with React, Next.js, TypeScript, Python, FastAPI, RESTful APIs, PostgreSQL, SQL Server, Redis, and ChromaDB, enhancing system reliability and reducing production incidents. Built and deployed cloud-native backend services on AWS with Docker, Kubernetes, Terraform, and CI/CD pipelines, increasing release frequency by 25% and improving deployment consistency. Streamlined database queries and introduced Redis caching strategies that reduced end-user response times by 15%. Integrated generative AI capabilities with OpenAI, Amazon Bedrock, LangChain, embeddings, ChromaDB, and vector search to return more relevant results and faster data retrieval, and designed RAG workflows combining document ingestion, embeddings, vector retrieval, and LLM-generated responses so users could get answers from enterprise knowledge sources with less wait time. Created Python and FastAPI services connecting enterprise data sources with AI search and knowledge management workflows, reducing business information retrieval time by 30%.

Developed full-stack web applications with React and JavaScript on the frontend and C#, GraphQL, and ASP.NET Core on the backend, shipping scalable features that improved clinician workflow in production healthcare applications. Established real-time communication with ASP.NET Core SignalR for responsive, event-driven updates between clients and backend services, reducing latency by 10%. Enhanced database performance by analyzing inefficient queries, improving data-access patterns, and tuning backend operations to cut response times by 30%, and integrated GraphQL APIs with C# services to streamline data retrieval and remove unnecessary frontend requests. Profiled the frontend, backend, and database layers with Visual Studio diagnostics to locate bottlenecks and apply targeted code and query optimizations that lowered overall latency. Partnered with cross-functional Scrum teams to design, test, and deliver over ten production features using Node.js, FastAPI, and DynamoDB, performing code reviews and monitoring performance with Prometheus to keep application stability high and avoid major outages.

Constructed enterprise applications with Angular, AngularJS, TypeScript, ASP.NET Core, MySQL, and Redis across the frontend, backend, and integration layers, improving performance by 20%. Built provider-search capabilities covering language, CME, location, and insurance coverage using Angular, ASP.NET Core, and Redis caching to improve search reliability. Refactored 36 VSource lookup models to support searches by last name and NPI while improving consistency across services, and tuned SQL queries and backend service calls to improve provider data retrieval and application responsiveness by 5%. Supported production applications used by 5,000+ users, quickly resolving frontend, backend, database, API, and integration issues while maintaining 99.9% system uptime. Conducted code reviews, automated testing, Maven builds, and production support with Git and JIRA to maintain code quality and reduce production incidents.