- Build and maintain microservices in C#/.NET on Docker, Azure, and Kubernetes.
- Own solution architecture, third-party integrations, integration tests, and build/deploy pipelines for multiple services.
- Led backend work on a multi-model assistant for client sites, including request routing between smaller and larger LLMs, a custom search component, and RAG over client content, with unit and integration tests (TestContainers) and CI/CD to Azure/Kubernetes.
20 years of backend and platform engineering, mostly in .NET, across a genuinely wide spread of industries — healthcare, industrial robotics, energy and the smart grid, market research, and IT consulting. The common thread across almost all of it is the same: solution architecture, third-party integrations, and service design for data-heavy products, usually as the person figuring out how the pieces fit together rather than just writing one piece of it. In the last few years that's extended into AI-driven production features (multi-model routing, RAG, LLM tooling folded into day-to-day development) and a deliberate effort to get good at directing AI tools rather than just using them.
Outside of day jobs, I ship a lot on my own — some of it serious, some of it just for fun. BillDesktop is a fully-featured SaaS billing platform I built and would happily sell. Nudge is a genuinely complex Android app with a real behavior-science-driven notification engine, live on Google Play with real users. Past that there's a long tail of smaller things: an Alexa-integrated AC remote, a handful of Android utilities and widgets, some crypto and sports-betting analysis tools that are mostly just fun excuses to build with data and models, and a home lab running real infrastructure. If you want the full picture of 20 years of shipping software, this page is where it lives — the one-page resume is the summary of the summary.
- Worked across multiple client assignments, including a React/MVC/Cosmos DB/Azure inventory platform.
- Consulted on .NET modernization efforts and built in-house tools for clients in .NET, AWS, and Postgres.
- Owned 20+ microservices and libraries across MVC/ASP.NET, .NET Core, Docker, CircleCI, AWS, MS SQL, Postgres, MQTT, CI/CD, Agile, and DI/IoC.
- Built automated simulation tests, data warehousing services, and ETLs.
- Worked across backend services, integrations, data workflows, customer portal features, and analytics.
- Implemented Alexa, Google, Ecobee, and Honeywell third-party integrations, plus internal tools such as Slack bots.
- Built the initial architecture and web presence for ZDT, including IoT reporting, alerts, and predictive monitoring.
- Built a survey generation platform to run complex, paid market-research surveys for clients.
- Built a community research platform — essentially message boards purpose-built for market research — with moderation, transcription, video encoding for multiple formats, polls, and incentive management, used to run long-running research communities for clients.
- Developed and maintained client-facing websites, primarily for GM.
- Built a CMS and introduced MS SQL-backed data storage for the team.
- Wrote VBA-based office automation tools to streamline workflows for the test lab department.
- A white-label customer billing platform: each business gets its own branded portal (logo, colors, own domain), invoicing that can be one-off or fully automatic recurring, and every common way a small business actually gets paid — Stripe card payments, ACH, Venmo/Zelle/Cash App logged and tracked, cash and check, partial payments and deposits. Beyond billing it handles scheduling and appointment requests, recurring service plans and memberships, customer self-service accounts (email, magic-link, or Google sign-in, optional 2FA), and an admin dashboard with a full audit log. Built and hosted on AWS with PostgreSQL and Stripe, wrapped in .NET.
- Rolling-average/std-dev models surface high-consistency candidates across NBA/NHL/MLB, backed by a Python analysis layer (backtesting, weather-adjusted modeling, statistical validation) on top of full-season enriched stat pipelines pulled from several public sports-stats APIs. Generates written analysis reports from the model output, plus a hosted Vue 3 dashboard for browsing results day-to-day.
- A video hosting platform for wrestling match footage — comments, timestamped notes on specific moments, playlists, event schedules, and a library of creator/wrestler profiles. Not in wide use yet, but fully built and running.
- Discovers ECS services and Lambdas in an AWS account and streams their CloudWatch logs live over a WebSocket, viewable from a local web UI.
- A personal home lab, not a single box: an old Acer gaming laptop repurposed as an Ubuntu Server running k3s, public-facing at a domain I own — hosting a dev mirror of this site, BetTracker, Gambler's dashboard, a self-hosted search engine, and a self-hosted Open WebUI. That Open WebUI sits on top of a private knowledge base built from personal notes, health data, and nightly-backed-up Telegram/text messages, so I can query my own data without sending it anywhere external. A separate Mac Mini on the internal network runs local LLMs via Ollama (including a finance-tuned model) and Home Assistant for smart-home control. DealTracker (github.com/dbraddillon/DealTracker), my public price-tracking project, runs here too — one of several small, genuinely useful things I've built just for the fun of automating my own life.
- Always-on .NET console app that lets you remote-control your dev box and Claude CLI from your phone via Telegram.
- A companion site to Gambler: screenshot a bet slip after the fact, Claude Vision parses the outcome, and it calculates a running win/loss record and P/L so I do not have to track it by hand.
- Windows Service that runs scheduled file-cleanup rules — filters files per folder by age and extension, deletes matches, and logs deleted/skipped/error counts per run.
- A local-first, single-user CRM I built to run my own job search end-to-end: company and contact tracking, an interaction timeline, follow-up tasks, resume version history, and a post-offer onboarding checklist system. It's also the source of this super-resume itself — a review workflow for turning a GitHub history into vetted, written-up content, which is what produced the page you're reading. Built deliberately outside my usual .NET/Vue stack (TypeScript/React/Node/Drizzle/SQLite) as a hands-on way to get fluent in that ecosystem.
- A layered Claude Code configuration, not just a config file: 9 custom subagents (dedicated architecture, security, performance, compliance, and test-coverage reviewers, plus separate senior-engineer review personas for C#/.NET, Go, and Java APIs), 13 custom slash-command skills (feature-spec writers, an implementation planner, a tradeoff-analysis facilitator, a repo bootstrapper, a wrap-up/handoff generator, among others), hooks that tie Claude Code into physical hardware (a USB-serial link that makes a desk gadget react to CLI activity) and guard against destructive commands (an ADB safety hook for Android work), and a managed plugin/marketplace layer. Composed once, reused across every other repo in this list. This is the actual argument for "I use AI tools well" — not the fact that I use them, but what I built to direct them.
- A .NET MAUI Android app for storing encrypted secrets — API keys, passwords, recovery codes — locally on-device, with a dedicated crypto service handling encryption at rest.
- A shared monorepo for Android home-screen widgets, each published as its own Play Store app from a common MAUI/AppWidgetProvider template — Sun Watch is the live one, with more in the pipeline.
- Combines a KenPom-style efficiency model (adjusted offense/defense/tempo data) with Claude-driven "buzz" analysis — a web-search tool-use call per team, blended into the scoring — to generate a full bracket as a structured, ready-to-enter pick sheet.
Additional side projects include smart-home tooling, an Alexa-integrated AC remote app, a Solana meme-coin tracking app, and a multi-chain explorer built to help with crypto tax reporting.
Android apps are all live on the Google Play Store; an iOS release is planned but not yet shipped.