Warrick Rodgers

Contact

Resume

SUMMARY

AI Software Engineer with 6+ years of experience, currently building Python-based LLM agent systems and document processing pipelines at Oracle. Daily work spans LLM agent tooling, agentic workflow development, and document ingestion pipelines, including work on a prior authorization automation system demoed at the executive level at Oracle. Proven track record shipping high-impact systems with executive visibility and tight delivery SLAs. Builds AI systems outside work too: architected DeskCurator, a four-agent autonomous content pipeline (TypeScript, ChromaDB, Gemini, SQLite), and WorkBuddy, a full-stack AI leadership coaching platform (React, PostgreSQL, RAG).

EXPERIENCE

Oracle

Kansas City, MO

IC2 Software Developer — Reimbursement AI Team Feb 2025 – Present

  • Contributed to agentic workflow tooling and developed a proof-of-concept UI for a prior authorization automation system, reducing a multi-day manual workflow to hours in testing, as part of a project that was demoed at the executive level, including to Oracle’s CEO.
  • Built document fetching and preprocessing tooling for an AI agent targeting healthcare claim denial workflows, enabling structured document context to be reliably passed to the LLM pipeline.
  • Designed and implemented fault-tolerant document ingestion and processing pipelines for LLM-driven agentic systems, with reliability and context quality as primary design constraints for a pre-release AI-powered product.
  • Collaborated cross-functionally with product, design, and AI research teams to rapidly prototype and iterate on agentic workflow tooling, accelerating delivery of executive-level product demos.

IC1 Software DeveloperOct 2022 – Jan 2025

  • Led RCA investigations into monitoring telemetry and system stability issues, consistently driving cross-team resolution within tight SLA windows.
  • Supported 5+ enterprise go-lives serving 100K+ end users; maintained sub-7-day detection-to-delivery SLA with 0 missed critical deadlines.
  • Resolved 10+ high-severity defects with executive visibility; 90% delivered within 72 hours, directly preventing downstream SLA breaches and client escalations.
  • Engaged directly with enterprise clients to clarify cloud-based release processes, producing updated documentation that improved release clarity for Oracle cloud solutions.
  • Facilitated weekly cross-domain knowledge transfer sessions across multiple functional teams, strengthening team-wide coverage and reducing escalations.

Cerner Corporation (Acquired by Oracle Corporation)

Kansas City, MO

Software Engineer IIFeb 2022 – Oct 2022

  • Collaborated with engineers across multiple countries to identify performance and design improvements on the Charge Services RESTful API.
  • Served as team subject matter expert for the Charge Services API, mentoring engineers and resolving workflow blockers.
  • Contributed to designs and implementations for automation scripts within Cerner’s RevElate project in the Patient Accounting organization.

Software Engineer IJan 2019 – Feb 2022

  • Co-designed and developed the Charge Services RESTful API V2, including GET endpoints and a proprietary POST endpoint; the API now processes 200K+ charges daily across integrated client workflows, built with test-driven development (TDD).
  • Identified and implemented GitHub workflow improvements, branch policies and PR standards, that reduced merge conflicts and improved deployment consistency across the team.
  • Built a Postman collection for the Charge Services API adopted by 10+ engineers, reducing data setup and testing time from 1–2 hours to minutes and decreasing support requests to the core team.

TECHNICAL SKILLS

Languages: Python (daily use, LLM/agent tooling, document processing, APIs), JavaScript (4 yrs), TypeScript (3 yrs), Java (2.5 yrs), SQL (5.5 yrs), C# (2 yrs), Ruby (1 yr), Dart (1 yr)

Frameworks & SDKs: React.js, Ruby on Rails, Flutter, Maven, OAuth 2.0, Pydantic, JUnit / RSpec / Jest

AI / LLM: LLM agent development, document ingestion pipelines, agentic workflow tooling, RAG, vector embeddings, prompt engineering, prior authorization automation

Cloud & Infra: OCI (Oracle Cloud Infrastructure), Docker, GitHub Actions, Jenkins (CI/CD)

Tools: GitHub, Jira, Postman, VSCode, IntelliJ, Visual Studio, VirtualBox

Engineering: RESTful API design, TDD, system design, full-stack development, Linux, observability & structured logging, web security, agile/scrum

PERSONAL PROJECTS

DeskCurator  ·  TypeScript · Node.js · ChromaDB · SQLite · Docker · Gemini API · WordPress REST API · Buffer API · Jest

Git: warrickrodgers/deskcurator

Four-agent AI content pipeline  ·  Phases 0–5a complete (~85%), social media agent in progress  ·  Mar 2026

  • Architected a four-agent autonomous system (ContentResearcher, ContentWriter, SeoOptimizer, SocialMediaAgent) with a two-phase research pipeline: AI-only product discovery followed by parallel Tavily-powered deep research with a 75% confidence threshold gate before any human approval is requested.
  • Designed and implemented a deterministic SEO scoring engine (9 pure-function checks, 100-point scale) with auto-fail conditions, AI-powered keyword validation, and an automated revision loop, articles revise and re-score up to 2 times before escalating to manual review.
  • Built a vector deduplication layer (ChromaDB, cosine similarity 0.85 threshold), a multi-provider AI abstraction service (Gemini + Anthropic Claude) with token bucket rate limiting, and differentiated error handling for TPM, RPD, and 503 failure classes.
  • Implemented a WordPress REST API publishing pipeline with SEO meta extraction, keyword-based category mapping, human-in-the-loop revision loops (up to 3 rounds), and a wp_failed recovery flow enabling credential-safe retries without re-generating content.
  • Built a SocialMediaAgent that triggers post-publish or on-demand via Discord command,  uses AI to decompose published articles into Twitter thread format, generates platform-optimized copy, and schedules posts to Buffer via API for optimal visibility timing.

WorkBuddy  ·  TypeScript · React · Node.js · PostgreSQL · Prisma · ChromaDB · Gemini API · JWT · Vite

Git: warrickrodgers/work_buddy

AI-powered leadership coaching platform  ·  In active development

  • Designed a three-service monorepo architecture (React/Vite frontend, Express orchestration layer, dedicated AI agent server) with strict service isolation, the frontend never contacts the AI service directly.
  • Built an AI agent (Simon) backed by a ChromaDB vector store seeded with a curated leadership knowledge base, enabling RAG-ready semantically grounded responses across 9 behavioral frameworks.
  • Implemented JWT authentication, paginated conversation persistence via PostgreSQL + Prisma ORM, and a multipart file upload pipeline feeding a structured data analysis intake workflow.
  • Engineered graceful degradation across service boundaries,  agent-server unavailability handled at the orchestration layer, preserving conversation state without data loss.

EDUCATION

Emporia State University2018

Bachelor of Science, Computer Science  ·  GPA: 3.53  ·  Cum Laude

Minor in Mathematics & Information Systems  ·  Phi Eta Sigma Honor Society

Bachelor of Arts, Psychology  ·  GPA: 3.53  ·  Cum Laude  ·  Psi Chi Honor Society