Things I’ve built trying to make an impact on the web and AI.

Here is a curated showcase of my engineering work in Generative AI, RAG architectures, agentic systems, and Responsible AI standards.

(Select up to 2 categories)
  • Very Hungry Tourists web application interface displaying food recommendations and grounded travel itineraries
    WebAI & Agentic Systems

    Very Hungry Tourists

    An AI-powered travel concierge and food trip planner built with Nuxt 3 and Gemini, using live Google Search grounding to curate real, operating local venues and sequence custom, budget-aware itineraries.

    Nuxt 3Gemini APISearch GroundingVercel
    Case Study
  • Parlez Bien app dashboard showing interactive AI language learning scenarios and student progress notes
    WebAI & Agentic Systems

    Parlez Bien

    A personalized AI language learning web app (Next.js & Vercel AI SDK) that generates tailored speaking scenarios from student notes, offering real-time tutoring with strict prompt-level guardrails for student confidence.

    Next.jsVercel AI SDKTypeScriptTailwind CSS
    Case Study
  • GRC Agentic AI System technical architecture diagram detailing metadata boundaries and multi-agent compliance retrieval
    AI & Agentic SystemsResearch & Standards

    GRC Agentic AI System

    An autonomous compliance reasoning platform leveraging agentic RAG (LangChain) to answer GDPR and ISO 27001 queries with physical metadata boundaries that prevent fabrication and ensure auditability.

    LangChainAgentic RAGPythonVector DBISO 27001
    Case Study
  • WebMCP Agent architecture visualization illustrating tool-calling pipelines and Model Context Protocol data streams
    AI & Agentic SystemsTools & Open Source

    WebMCP Filter Agent

    An open-source agentic web tool leveraging the Model Context Protocol (MCP) to scrape, filter, and extract actionable meeting minutes and key decisions with structured tool-calling pipelines.

    WebMCPModel Context ProtocolTypeScriptNode.js
    View GitHub Repo
  • Search Grounding Benchmark evaluation harness architecture and automated test flow diagram
    Research & StandardsTools & Open Source

    Search Grounding Benchmark

    A reproducible test harness and benchmark suite auditing LLM search grounding factual failure rates, zombie data retention, and locality drift across 30 queries in 10 global cities.

    TypeScriptGemini REST APIBenchmarkingOpen Source
    View GitHub Repo