I enjoy exploring the intersection of AI and UX — combining fullstack engineering, design thinking, and AI research to build products that make complex technology accessible and empowering.
Rabbit Hole — Real-time collaborative knowledge graph platform
Built sync infrastructure (Yjs CRDT, <100ms latency), integrated LangGraph agents for automated research, and designed tier system handling Free through Enterprise users.
Stack: React, Next.js, TypeScript, Node, Neo4j, PostgreSQL
Deep Agent Researcher — Multi-agent knowledge extraction system
Supervisor coordinates 6 specialized agents using LangGraph for autonomous research workflows. Produces validated knowledge graph bundles with evidence citations and confidence scores.
Stack: LangGraph, TypeScript, Python
AI Playground System — Interactive capability testing framework
Reusable component pattern for validating AI features before full implementation. Built 12 playgrounds for LLM providers, transcription, and agent workflows. Enables non-technical stakeholders to test directly.
Stack: React Query, TypeScript, Storybook
- acai.so — AI-powered markdown editor with agentic workflows and local model support
- AC-Scripts — Open-source AI scripts for developer productivity (ScriptKit ecosystem)
- Chat Your Knowledgebase — Primer on building RAG pipelines with LLMs & embeddings
Frontend: React, Next.js 15, component libraries, design systems, accessibility, performance optimization
Backend: Node.js, Python, API design, WebSockets, Docker, monorepo architecture
Data: PostgreSQL, Neo4j, Redis, vector databases, S3
AI/ML: LangChain, LangGraph, prompt engineering, RAG pipelines, tool orchestration (OpenAI, Anthropic, Groq)
Infrastructure: Real-time collaboration (Yjs CRDT), CI/CD, testing (Vitest, Cypress), observability
I believe AI should augment developer creativity, not replace it. My focus is on building tools that enhance productivity while keeping safety, transparency, and maintainability at the core. Learn by shipping, measure what matters, iterate.