Experience
Production platforms, client-facing quality.
SolutionStream
AI Engineer — client-facing product
Current
Client work stays confidential here — what follows is the shape of the engineering, not its details.
I work on a client-facing product platform as an AI engineer: building AI-driven features and the full-stack product around them, and owning the quality work that keeps a production system trustworthy — route behavior, data coverage, error states, regression testing, and release validation.
AI product features
Building AI-driven capabilities into a client-facing product — and the systems thinking that keeps them trustworthy: what the model may touch, what stays deterministic, and how results are validated.
Product engineering
Client-facing features across the stack: React and Next.js frontends, API endpoints, and PostgreSQL-backed data systems — implemented end to end, from route behavior to the data underneath it.
Data coverage and honest states
A recurring theme of my platform work: making sure routes behave correctly when data is partial, unavailable, or failing — and that empty, error, and loading states tell the truth instead of pretending.
Quality and release validation
Regression testing with Playwright, CI and release-gate validation, and pre-merge verification. The habit is the same one my personal projects show: evidence before a status claim.
Production investigation
Debugging live behavior, auditing architecture and unfamiliar areas of the codebase, and writing up findings so the next engineer doesn’t re-derive them.
Engineering process
Issue-driven work, reviewable pull requests, and technical documentation. I write things down — investigation reports, decision records, runbooks for what I touched.
AI-assisted engineering
Using AI tooling to move faster while keeping the verification human and deterministic: generated code is reviewed, tested, and validated against evidence before it counts as done.
Southern Utah University
Research Assistant
+ Open source
Contributing to codebases I don’t own.
My own repositories show how I build from scratch; contribution shows the other skill — reading an unfamiliar codebase, following its conventions, and making a change its maintainers can trust. My projects are open source — MIT and Apache-2.0 — with contribution guides, and RepoSignal ships good-first-issues scoped for new contributors.
+ The stack
Multiple stacks, one standard.
Every language here appears in a shipped, public project or in production work — no proficiency bars, just where things get used. The engineering standard travels across all of them.
Languages
- TypeScript
- JavaScript
- Python
- Go
- Rust
- C#
- Swift
- SQL
- C
- C++
Frontend
- React
- Next.js
- Svelte
- Astro
- HTMX
- SwiftUI
- Tailwind CSS
- HTML + CSS
Backend
- Node.js
- FastAPI
- ASP.NET Core
- Go services
- REST APIs
- GraphQL
Data
- PostgreSQL
- SQLite
- Prisma
- SQLAlchemy
- Entity Framework
- Supabase
Testing & delivery
- Playwright
- Vitest
- pytest
- xUnit
- GitHub Actions
- CI/CD
AI & tooling
- RAG
- semantic retrieval
- AI agents
- context engineering
- Ollama
- Git + GitHub
The full picture, projects included, lives on one page.