Software Platform Architect / Founder — MyCritters
2022-07 - Present
Pewaukee, Wisconsin
- Led product and engineering delivery as a 100% remote founder-operator and team manager.
- Architected and shipped a full-stack platform using Next.js, TypeScript, GraphQL, serverless functions, and PostgreSQL/Hasura.
- Evaluated tRPC for typed API workflows during early platform iterations, then consolidated on GraphQL to simplify contracts and role-based data access.
- Built a Go API server with GraphQL interfaces to integrate third-party Avimark practice-management data and workflows into the platform.
- Built event-driven pipelines with BullMQ (Redis-backed queues) and asynchronous workers for scraping, enrichment, and product workflows.
- Implemented multi-role RBAC (Hasura roles, Nhost auth, serverless auth functions) for facilities, staff, and client accounts.
- Processed over 500k webpages in production scraping pipelines with AI-assisted enrichment and structured outputs.
- Implemented hybrid facility discovery by running AI embedding + PostgreSQL search in parallel with Typesense, then merging and ranking results for user-facing relevance.
- Deployed and operated containerized Typesense infrastructure with secured cron-driven multi-collection cleanup and sync workflows.
- Integrated AI capabilities into both product features and team delivery workflows to speed iteration and improve implementation quality.
- Instrumented MyCritters site and product funnels with Google Analytics (GA4), Google Search Console, and PostHog to monitor acquisition, engagement, and conversion behavior.
- Built GraphQL WebSocket subscription support with lazy connection initialization and reconnect/backoff handling for efficient real-time updates.
- Selected AI/ML Systems:
- Built a production RAG pipeline using graph + embeddings retrieval for domain knowledge search and answer generation.
- Evaluated structural and semantic chunking strategies and implemented embedding reprocessing/versioning to improve retrieval quality over time.
- Designed an entity extraction pipeline to convert unstructured web content into normalized, structured records.
- Deployed and operated multiple containerized LLM inference services on Kubernetes with GPU/CPU profiles and performance tuning.
- Designed and deployed an end-to-end Python (Flask) speech-to-text + diarization microservice (Whisper + Pyannote) in Docker, supporting GPU acceleration, resource-constrained CPU execution, and upload/record web UI workflows across local and cluster environments.
- Internship Program Leadership:
- Mentored 28 interns through the program, with up to 12 participating concurrently, in a remote development environment.
- Trained 4 lead engineers who managed distributed development teams and day-to-day execution.
- Deployed a fully cloud-based development platform with approximately 1-hour onboarding, enabling new developers to write production code on day one.
- Built onboarding flows, technical curriculum, code review standards, and CI/CD templates that enabled contributors to ship production code quickly.
- Led delivery planning across up to 7 concurrent initiatives, coordinating milestones, quality, and execution.