AI & software architect · Hands-on engineer · Kelowna, BC, Canada
I turn complex product ideas into practical systems, from early discovery and prototypes through architecture, security review, and production delivery. My 18+ years in software span retail innovation, healthcare AI, mobile products, and game operations.
- Applied AI: agentic systems, retrieval, knowledge graphs, and model cost and reliability trade-offs.
- Backend and cloud platforms: APIs, distributed systems, data pipelines, and Azure and AWS architecture.
- Technical leadership: connecting product goals with engineering decisions and staying involved in implementation.
Core tools: Python, PostgreSQL, Azure AI Search, Redis, Azure, and AWS.
My public contributions focus on reproducible bugs, focused fixes, and regression tests. These pull requests were submitted in October 2026 and are awaiting review; each link shows its current status.
| Project | Contribution |
|---|---|
| Graphiti | Handle empty node-label filters in knowledge-graph search. |
| FastEmbed | Avoid unnecessary backoff after the final failed model download. |
| Firecrawl | Preserve nested crawl-preview settings in the async Python SDK. |
| Browser Use | Honor zero pricing-cache retention while preserving other sources' files. |
| Dify | Verify package-icon persistence and ownership with real SQLite-backed tests. |
| AutoGPT | Preserve exact large integers when converting text-based workflow inputs. |
| Hermes Agent | Keep invalid persisted scheduling settings on the documented fallback path. |
The contribution forks on this profile belong to their respective upstream projects. The links above point to my submitted changes.
Much of my product work lives in private repositories. This profile highlights the work I can share publicly.