Forward Deployed AI Engineer · AI Solutions Architect · Founder, wovv.ai
I sit with clients, understand their process and real pain points, design the right AI solution, and own delivery end to end — until it works in production. Not advice handed off, but systems shipped.
I bring a combination most engineers have only one piece of:
- Field-deployed delivery — a decade taking software into real-world operations: digitizing Walmart's 10,000+ driver fleet, on-site AR for BMW technicians, and pharmacy at Albertsons scale.
- Hands-on GenAI depth — RAG, multi-agent systems, knowledge graphs, and LLM evals, in production rather than in slides.
- Engineering ownership — I take a solution from client need to working system, and I'm accountable for the outcome.
🛠️ Currently building wovv.ai
An AI-native storytelling platform — a personal AI writing partner for novelists and screenwriters. I design the product and the multi-agent engineering system that ships it:
- 3-service platform from scratch: Next.js web app, FastAPI backend with RAG + knowledge-graph memory (Postgres + Neo4j), and a GPU media-generation service.
- The engineering org itself designed as a multi-agent system — AI "councils" that plan, build, review, and govern under formal decision records.
- A golden-set eval + canary-invariant quality system validating LLM output before every release.
- All cloud infra as code on GCP (Pulumi, Cloud Run incl. GPU, scale-to-zero), held under $700/month at full capacity.
Most of my recent build work lives in private repos. The clearest look at what I build is the live product at wovv.ai — and I'm always happy to do a screen-share walkthrough of the architecture and code.
AI/GenAI: LLMs · RAG · Knowledge Graphs · Multi-Agent Systems · LLM Evals · LangChain Backend & Cloud: Python · FastAPI · Next.js · GCP · Cloud Run · Pulumi (IaC) · Postgres/pgvector · Neo4j Mobile & AR: Android · Kotlin · Jetpack Compose · ARCore · Azure Spatial Anchors · JNI