Senior full-stack software engineer with 10+ years of experience shipping production web applications, backend services, APIs, internal tools, and enterprise platforms. Recent work focuses on AI-enabled product engineering: LLM workflows, RAG prototypes, embeddings, graph-based retrieval, tool calling, agentic systems, AI-assisted development, and model evaluation. Built and shipped WingIt as solo founder, owning product planning, architecture, UI, backend APIs, CI/CD, container delivery, analytics, and production observability. Strong hands-on background across Python, FastAPI, TypeScript, React, Next.js, C#, .NET, Docker, Azure, AWS, SQL, Redis, Neo4j, and production debugging. Comfortable in startup and small-team environments where senior engineers are expected to move from ambiguous product ideas to deployed software.
Experience
Built and shipped WingIt, a production full-stack AI web application, owning every layer from product spec to deployment.
- Designed, built, and deployed WingIt end to end, including architecture, data model, UI, backend APIs, CI/CD, container delivery, analytics, and production observability.
- Built a full-stack product using Next.js, TypeScript, Tailwind CSS, backend APIs, Docker-based delivery, GitHub Actions, and GHCR.
- Instrumented logs, metrics, traces, session replay, funnels, and conversion dashboards to guide product iteration using real behavior data.
- Used Claude, Codex, ChatGPT, and Kiro heavily across the build for planning, code generation, refactoring, debugging, and architectural critique.
- Authored product specs, implementation plans, and acceptance criteria before prompting AI coding tools, then cross-critiqued model outputs to catch ambiguity, brittle assumptions, and failure modes.
- Worked in a startup-style environment with full ownership of product direction, technical execution, deployment, and iteration.
Python-first contractor building evaluation harnesses, RAG-style workflows, and agentic pipelines to generate high-signal data for evaluating and improving code-capable language models.
- Built FastAPI microservices and Pydantic/SQLAlchemy data layers for speech-to-text transcription, audio segmentation, and feature extraction using Whisper and Librosa.
- Prototyped a Neo4j-backed RAG system using embeddings, graph retrieval, structured logging, and domain-specific context expansion for audio-related knowledge.
- Orchestrated Dockerized services in an agentic platform performing coordinated tool calls, task execution, and reusable workflow steps.
- Helped refine agentic patterns for planning, tool selection, execution quality, and reusable-task design.
- Characterized recurring model failure modes and translated them into authored tasks, grading rubrics, and evaluation criteria for downstream model improvement.
Embedded contractor working directly with a frontier AI lab on software-engineering capability improvements for coding models.
- Completed onboarding and was immediately selected for the accelerated experts track.
- Promoted to occupational reviewer for software development and AI model training, judging other contributors' tasks, rubrics, and technical reasoning.
- Reviewed software engineering tasks, code diffs, model outputs, and evaluation rubrics for correctness, maintainability, and signal quality.
- Set technical direction on critical projects and enforced Python and TypeScript coding standards across cross-functional contributors.
- Mentored junior contributors on code review expectations, task clarity, rubric design, and stronger engineering judgment.
Benchmarked and improved top LLMs on software-engineering capabilities, including code generation, debugging, architecture, deployment, and technical explanation.
- Authored and reviewed RLHF tasks across code generation, bug detection, debugging, performance, and parallelism in Python, C++, JavaScript, and TypeScript.
- Reviewed model transcripts and contributor submissions, writing defensible technical reasoning explaining why one implementation wins, where tests are weak, and what a diff actually changes.
- Deployed React, Next.js, Tailwind CSS, and TypeScript frontends in sandboxed agentic workflows to produce realistic, executable evaluation environments.
- Investigated and debugged large enterprise-style codebases with multiple services, APIs, web-service integrations, and deployment workflows.
- Designed CI/CD and DevOps workflows on AWS with Docker and Portainer for model evaluation tasks and sandboxed application environments.
- Promoted to senior reviewer on multiple projects and helped onboard junior contributors to AI software evaluation workflows.
- Contributed to advanced AI benchmark content, including Humanity's Last Exam, with focus on technical correctness, rubric quality, reasoning quality, and high-signal evaluation design.
Technical lead on a platform of .NET web applications, REST APIs, SOAP services, identity systems, and enterprise tools serving major airlines, GDSs, and travel platforms.
- Designed, built, and maintained .NET applications and REST APIs used by Lufthansa, American Airlines, United Airlines, Delta Air Lines, Travelport, Priceline, and other enterprise customers.
- Designed and shipped an OIDC/OAuth 2.0 identity and access management system using ASP.NET Core and IdentityServer 4 for federated SSO across airline customers.
- Owned a critical C++/.NET SOAP messaging gateway responsible for dispatching platform traffic across enterprise services.
- Built OpenAPI and Swagger-documented APIs for enterprise integrations and internal administrative tools.
- Implemented a SAX-based XML validator that materially improved transaction consistency across high-volume travel workflows.
- Built a self-service administrative portal that allowed airline customers to manage tenants and onboard travel agents, reducing onboarding time from weeks to minutes.
- Designed a Redis caching layer to improve data retrieval performance under load.
- Implemented Serilog sinks and contributed to centralized logging, metrics, and observability workflows using Greylog, DataDog, and platform telemetry.
- Promoted to technical lead overseeing a suite of customer-facing .NET applications, including staging and development environment design, CI/CD pipelines, IIS configuration, certificates, firewall rules, load balancers, and VIPs.
Software engineer in the Reliability Department building Python, C++, and Java software for hard drive testing, acoustic analysis, internal tooling, logging, and CI/CD.
- Built Python reliability test suites enabling early detection of serious hard drive manufacturing and reliability defects.
- Authored and maintained C++ libraries with Boost.Log and Google Test, providing platform-wide logging, unit testing, and mocking primitives.
- Developed Windows Forms applications for acoustic testing of hard disk drives.
- Set department-wide coding standards and automated API documentation generation from source code comments.
- Deployed Atlassian Bamboo CI/CD across the team to improve build, test, and deployment efficiency.
Led backend web development for the College of Business web platform while completing a computer science degree.
- Led backend web development on the college web platform using C#, ASP.NET, SharePoint Web Parts, AJAX-enhanced web forms, and REST service integrations.
- Prototyped WCF services and internal tools to make platform data available across applications.
- Promoted to Team Lead managing a small development team while concurrently starting an internship at Seagate.
Projects
https://wingit.dev | Production full-stack AI presentation app that turns live speech into generated slides, captions, and audience views in real time. Built solo from product concept through deployment, including architecture, data model, UI, backend APIs, Docker-based delivery, CI/CD, analytics, and production observability.
https://rolledex.netlify.app | Event networking intelligence app for meetups, hackathons, and professional events. Converts event context, attendee information, company research, and follow-up goals into event dossiers, people cards, company briefs, and practical networking actions.
https://sdn-openapi.netlify.app/ | Prototype combining a custom SDN list API for OFAC sanctions data with a Subjective Logic 209 reasoning project for uncertainty-aware analysis. Explores how structured compliance data, fuzzy entity matching, and subjective logic can support probabilistic reasoning about sanctions exposure, geopolitical signals, entity relationships, and downstream financial or operational risk.
https://root-note.netlify.app | AI DJ and guided music discovery app that turns a genre, artist, sample, scene, era, or music rabbit hole into a playable YouTube listening episode with host commentary, adaptive visual mood, traceable roots, and optional ElevenLabs narration.