Backend and platform engineer working on reliability boundaries, data systems, realtime systems, and evidence-governed AI-assisted engineering.

I build systems that make state, failure, evidence, and recovery inspectable—not merely systems that appear to work in a passing demo.

A Rust, local-first transcript QA system built around bounded evidence, deterministic review semantics, explicit human decisions, and reproducible validation.

The current public baseline establishes a bounded v0.1 core mechanism and persistence layer. Broader semantic and model-assisted behavior remains experimental rather than authoritative.

An exploratory Rust and TypeScript music-systems foundation for explicit musical state, deterministic semantic operations, transport-independent protocol boundaries, and performance measurement.

The earlier collaboration-led demo direction did not establish a product differentiator and is retained as a failed product-narrative experiment, not as a current collaborative-composition claim. The repository demonstrates bounded architecture and evidence work rather than a production-ready DAW or an accepted product thesis.

An inspectable, test-backed reference for reliability semantics at the database-to-message boundary.

It makes event identity, idempotent local effects, concurrent worker transitions, retry and dead-letter behavior, and the boundary between local transactions and external delivery explicit and testable. It also includes a deterministic red-to-green reproduction of a stale failure-write race.

I am available for short, bounded backend and platform engagements where correctness, operational risk, data flow, or system boundaries matter.

Typical work includes:

• backend reliability and failure-mode audits
• PostgreSQL schema, query, migration, and consistency diagnosis
• worker, queue, retry, idempotency, outbox, and inbox review
• legacy .NET/C# or TypeScript/Node.js backend triage
• realtime APIs, state synchronization, and internal tooling
• local-first or human-in-the-loop AI workflow prototypes

I prefer engagements with explicit scope, inspectable deliverables, written findings, and clearly stated tradeoffs.

C# · .NET · Go · TypeScript · Node.js · Rust · Python

PostgreSQL · TimescaleDB · PostGIS · MySQL · NATS JetStream · Kafka

Docker · GitHub Actions · Testcontainers · Prometheus · Grafana · Linux

React · Vite · WebSocket · Socket.IO

Projects here vary in maturity. Their READMEs aim to state current scope, guarantees, evidence, and limitations explicitly.