BRENT OAKES / CHICAGO, IL

Shipping AI is an engineering problem.

I build AI software for healthcare and enterprise teams, from agentic workflows and knowledge retrieval to real-time applications that people rely on.

See selected work Download résumé
PRODUCTION SYSTEM / 01HEALTHCARE AI
Representative interface for a clinical AI assistant platform
Architecture → Delivery → Evaluation✳
35%lower response
latency
ENGINEERING SOFTWARE SINCE 2019SCROLL TO EXPLORE ↓

Selected impact across
production systems

1M+documents supported
35%lower response latency
22%fewer unsupported responses

01 / SELECTED WORK

Systems with real stakes.

Four examples of engineering decisions translated into working products and measurable outcomes.

Representative clinical AI assistant interface
PROJECT / 01REPRESENTATIVE INTERFACE

01 / Healthcare AI platform

Clinical AI, built to operate at scale.

Led the architecture and delivery of an enterprise platform connecting agent workflows, retrieval, voice, and operational automation.

35%lower response latency

THE CHALLENGE

Clinical workflows required grounded answers, secure enterprise integrations, and responsive experiences for thousands of users.

WHAT I BUILT

Designed supervisor-based orchestration and specialized agents with LangGraph and MCP. Built asynchronous FastAPI services with streaming, caching, and evaluation.

Python / FastAPI / LangGraph / MCP / AWS

Representative AI model orchestration interface
PROJECT / 02REPRESENTATIVE INTERFACE

02 / Enterprise knowledge systems

Useful answers from a million documents.

Created a retrieval and model-routing platform that turns large clinical and enterprise knowledge sets into grounded AI responses.

22%fewer unsupported responses

THE CHALLENGE

Finding the right evidence across a large document collection called for more than basic vector search.

WHAT I BUILT

Built ingestion, hybrid retrieval, metadata filtering, reranking, and evaluation. Routed tasks across foundation models by capability, latency, and cost.

RAG / Vector search / OpenAI / Claude / Gemini / Bedrock

Representative conversational real estate platform interface
PROJECT / 03REPRESENTATIVE INTERFACE

03 / Property technology

A better way to search, speak, and discover.

Shipped conversational property search, listing generation, and voice features for a customer-facing SaaS product.

32%faster conversational response

THE CHALLENGE

The product needed to respond naturally across text and speech while keeping each interaction relevant to the customer.

WHAT I BUILT

Built RAG and real-time voice pipelines with FastAPI, React, and Next.js, supported by personalization and memory services.

Python / FastAPI / React / Next.js / AWS

Representative healthcare analytics dashboard
PROJECT / 04REPRESENTATIVE INTERFACE

04 / Healthcare software

The dependable software beneath the workflow.

Built healthcare applications, analytics, backend APIs, and secure access patterns for clinical and operational teams.

30%faster database queries

THE CHALLENGE

Everyday workflows needed reliable data access, fast queries, and clear permissions.

WHAT I BUILT

Delivered React and Node.js applications, service integrations, role-based authorization, and database tuning.

React / Node.js / Python / PostgreSQL / AWS

02 / EXPERIENCE

Built across the stack.

A progression from healthcare software to production AI architecture.

2023 — 2026

Senior AI Software Engineer

Elastech

Production healthcare AI platforms, agent orchestration, retrieval, evaluation, and cloud services.

2021 — 2023

AI Full Stack Engineer

Grayphite

Conversational search, voice experiences, personalization, and AI-powered SaaS delivery.

2019 — 2021

Full Stack Developer

Epic

Clinical workflows, analytics, APIs, secure data systems, and cloud applications.

View full résumé

03 / EXPERTISE

The right tools for the system.

I work across architecture, product engineering, and operations to get AI into users’ hands.

01 / AI SYSTEMS

Orchestrate

Agent workflows, model routing, RAG, voice AI, and grounded responses.

LANGGRAPH / MCP / OPENAI / BEDROCK
02 / PRODUCT ENGINEERING

Build

Reliable APIs, streaming services, and interfaces designed around actual workflows.

PYTHON / FASTAPI / TYPESCRIPT / REACT
03 / PRODUCTION

Operate

Evaluation, observability, security, latency, and cloud deployment from the start.

AWS / DOCKER / POSTGRESQL / REDIS

04 / ABOUT

The engineer behind the work.

I’m Brent, a Chicago-based engineer focused on making complex AI useful in demanding environments.

I bring 7+ years of software engineering experience, including healthcare applications, AI-powered SaaS, and enterprise AI platforms. I studied computer science at Trinity University.

My work is strongest where architecture meets product reality: understanding the workflow, choosing the right technical approach, and measuring how the system performs after launch.

05 / CONTACT

Let’s make the next
system count.

Open to senior AI engineering opportunities and thoughtful collaborations.

Email me brentoakes025@gmail.comLinkedIn Connect and messageGitHub View public repositoriesRésumé Download PDF