Brent Oakes
Designing and scaling LLM-powered platforms, agentic workflows, RAG systems, and real-time AI applications from architecture through deployment for healthcare and enterprise products.
About Me
I'm a Senior AI Software Engineer with 7+ years of software engineering experience and 5+ years building production AI systems for healthcare and enterprise applications.
My expertise spans agent orchestration and intelligent model routing, large-scale RAG systems, multimodal and real-time conversational AI, evaluation, observability, and cloud-native delivery. I build with Python, FastAPI, C#/.NET, TypeScript, React, and AWS, integrating OpenAI, Anthropic Claude, Google Gemini, AWS Bedrock, and open-source models into reliable products.
Based in Chicago, IL, I've delivered AI platforms supporting 1M+ documents and thousands of users, reducing unsupported AI responses by 22% and improving response latency by 35%.
Work Experience
Production AI systems, healthcare applications, and cloud-native platforms delivered from architecture through deployment.
Elastech
Senior AI Software Engineer
- Led the architecture and production delivery of an enterprise healthcare AI platform integrating agentic AI workflows, RAG systems, real-time conversational AI, and enterprise workflow automation for clinical and operational applications.
- Architected agent-based AI workflows using LangGraph, LangChain, and MCP with supervisor-based orchestration, specialized agents, workflow state management, tool calling, memory handling, and secure integrations with healthcare enterprise systems.
- Orchestrated intelligent LLM model routing within AI agents across OpenAI, Anthropic Claude, Google Gemini, and AWS Bedrock, selecting foundation models dynamically based on task complexity, latency requirements, capability, and cost optimization.
- Integrated text, voice, and document capabilities into AI workflows through real-time speech pipelines, document processing systems, retrieval components, and context-aware reasoning for healthcare use cases.
- Delivered an enterprise RAG platform supporting 1M+ clinical and enterprise documents through ingestion pipelines, embeddings, hybrid retrieval, vector search, metadata filtering, reranking, and retrieval evaluation, reducing unsupported AI responses by 22%.
- Optimized scalable AI backend services using Python, FastAPI, asynchronous processing, streaming APIs, caching strategies, and cloud-native deployment patterns, supporting thousands of users while improving response latency by 35%.
- Established production AI engineering practices covering LLM evaluation, agent evaluation, observability, hallucination detection, prompt regression testing, retrieval quality measurement, security controls, and reliable AI operations.
Grayphite
AI Full Stack Engineer
- Delivered AI-powered SaaS features including conversational property search, automated listing generation, and intelligent content workflows powered by LLM-based applications.
- Engineered end-to-end RAG applications using Python, FastAPI, React, Next.js, embeddings, and vector search pipelines for customer-facing products.
- Integrated LLM workflows with prompt engineering, structured outputs, retrieval strategies, and evaluation methods to improve response consistency and reliability.
- Implemented real-time conversational features combining speech recognition, LLM reasoning, and text-to-speech services, improving conversational responsiveness by approximately 32%.
- Built personalization and memory services using PostgreSQL, DynamoDB, Redis, and vector embeddings to support context-aware user experiences.
- Deployed cloud-native AI applications on AWS using Docker, ECS, Lambda, Terraform, and CI/CD pipelines while connecting AI capabilities with existing SaaS platforms.
Epic
Full Stack Developer
- Built enterprise healthcare applications, workflow systems, and analytics dashboards using React, JavaScript, Node.js, and Python for clinical and operational workflows.
- Designed backend APIs and service integrations supporting healthcare data processing, reporting workflows, automation, and distributed application functionality.
- Implemented secure application access using OAuth 2.0, JWT, and role-based authorization patterns for systems handling sensitive healthcare information.
- Improved PostgreSQL and MySQL performance through schema optimization, indexing strategies, and query tuning, reducing execution times by approximately 30%.
- Improved AWS-hosted applications using EC2, ECS, CloudFormation, and CodePipeline to increase deployment reliability and operational efficiency.
Featured Projects
Production systems that power enterprise AI capabilities and serve thousands of users.
Enterprise Healthcare Agentic AI Platform
Production healthcare AI platform combining agentic workflows, RAG, real-time conversational AI, and enterprise workflow automation.
Enterprise RAG & Model Routing Platform
Large-scale RAG and intelligent model-routing platform spanning more than one million clinical and enterprise documents.
Real Estate SaaS with Conversational AI
AI-powered customer engagement platform with conversational property search, automated listing generation, and real-time voice AI.
Healthcare Analytics & Workflow Platform
Enterprise healthcare applications, workflow systems, analytics dashboards, backend APIs, and secure cloud integrations.
Technical Skills
Deep expertise across the full AI and software engineering stack.
Agentic AI Engineering
LLM & Generative AI
AI Quality, Evaluation & Observability
RAG & Knowledge Systems
Backend & Full Stack Engineering
Cloud, Infrastructure & DevOps
Data & Security
Let's Build Together
I'm always interested in discussing new AI engineering challenges, enterprise platform opportunities, or innovative projects.
Send Me an EmailPrefer a different platform? Reach out via: