Ruby on Rails Backend Engineer · API Reliability & LLM Integrations

I'm a backend engineer with 5+ years of professional experience building and maintaining Ruby on Rails applications. I work with PostgreSQL-backed services, REST APIs, third-party integrations, and existing systems that need to stay reliable and maintainable.

My current focus is reliable AI integrations: predictable retries and timeouts, strict response validation, privacy-conscious observability, and testable provider boundaries.

A reproducible, local-only Rails engineering lab demonstrating how to handle unreliable LLM APIs safely. It uses synthetic support tickets and a deterministic Fake Provider by default; the browser dashboard never makes a paid LLM request.

• Resilience: bounded retries, Retry-After, typed failures, an 8-second per-attempt budget and a 25-second overall cooperative deadline.
• Provider design: eight repeatable failure scenarios; optional, explicitly enabled Anthropic Ruby SDK adapter with SDK retries disabled and independent JSON-schema validation.
• Privacy and observability: metadata-only PostgreSQL runs, allowlisted Rails structured events, Basic Auth, CSRF protection, and escaped browser output.
• Verification: 396 passing RSpec examples in the final October 2026 portfolio audit, with RuboCop and Brakeman checks passing in GitHub Actions.

Explore: Architecture · Simulated engineering audit · 3–5 minute walkthrough · CI

This is a synthetic engineering demonstration, not a production deployment or a real-world model-accuracy benchmark. Live Anthropic behavior has not been verified.

• Fitness Tracker — Rails workout tracking with Devise authentication, exercises, training sets, and charts.
• File Sorter — a desktop utility for organizing files.
• 20-20-20 Rule — a small Linux screen-break reminder script.
• Hieroglyphic Name — a creative web experiment inspired by Egyptian hieroglyphics.

• Backend: Ruby, Rails, REST APIs, PostgreSQL, application maintenance and modernization
• Reliability & integrations: retry/backoff policies, deadline management, Stripe, third-party APIs, LLM provider adapters
• Testing & tooling: RSpec, WebMock, GitHub Actions, RuboCop, Brakeman, Docker, Git, Linux
• AI development: LLM application architecture, Claude Code, OpenAI Codex, MCP

Claude Certified Architect — Foundations (CCA-F), Anthropic.

Outside software, I enjoy electronics, microcontrollers, guitars, and experimental builds.