I build agentic AI systems and the platforms they run on. Current work centers on multi-agent orchestration with Ruflo and retrieval infrastructure on RuVector. Behind that sits a decade of full-stack delivery for organizations such as Lenovo, UNICEF, Honda Marine, Carnegie Mellon University, and Weill Cornell.
- Multi-agent orchestration: Ruflo workflows coordinating specialized agents across research, implementation, review, and QA, with instrumented runs measuring token cost, latency, and task success
- Retrieval infrastructure: RAG and semantic search on RuVector and PostgreSQL, covering chunking strategy, embedding selection, and retrieval-quality evaluation
- MCP tooling: Model Context Protocol servers exposing data and operations to agent runtimes under controlled scopes
- Application delivery: Next.js and TypeScript front ends over Python and Node.js services, containerized and shipped through CI/CD
- Enterprise Drupal: complex migrations (D5 to D11), performance optimization (LCP 4.2s to 1.9s on a UNICEF platform migration), multilingual architectures, custom modules
- Headless and decoupled: API-first architectures bridging Drupal back ends with React and Next.js front ends
Looking for an expert to elevate your project? Let’s connect! 🚀