Signal is an interactive portfolio built as a multi-service system. Instead of a static page, it gives you a conversational way to explore my work, projects, and leadership experience. Under the hood, it demonstrates how I approach frontend craft, system design, and modern AI workflows.
Signal is a proof point. It shows how I design systems that balance clarity, maintainability, and developer experience, while experimenting with modern AI patterns like Retrieval-Augmented Generation (RAG) and Model Context Protocol (MCP).
It's less about shipping a product and more about demonstrating service boundaries, shared standards, and production-minded practices in a portfolio context.
- Frontend: Next.js 15, React 19, Tailwind CSS, TypeScript
- Backend: Node.js + Express
- Retrieval: ChromaDB (RAG) with OpenAI embeddings
- Integrations: Model Context Protocol (MCP) server for GitHub, Spotify, and blog
- Model orchestration: OpenRouter (Gemini Flash 2.5) with tool-calling enabled
- Dev tooling: ESLint, Prettier, Husky, and shared configs via dev-config
- Ops: Docker for portability, GitHub Actions + semantic-release for automation
- Hosting: Fly.io
Signal is both portfolio and proof point. It shows my frontend expertise through a polished, real-time chat interface. It demonstrates system-level thinking with clear service boundaries and shared standards. It reflects a leadership mindset by focusing on maintainable, documented systems that align with real-world production practices.
At a higher level, Signal reflects how I think about building systems that balance clarity, maintainability, and team enablement.