All Projects
Agentic AI systems across four patterns: ReAct, Plan-and-Execute, Multi-Agent, and Self-Correcting.
Tool-Usewip
Agent Skills Dashboard
Performance charts comparing GPT-4o vs. Llama 3 across my agents, tool-use accuracy metrics, and a live playground to trigger real agent actions.
RechartsPostgresOpenAILlama 3+2 more
ReActwip
Deep Research Agent
An autonomous agent that browses, fact-checks, and generates a PDF report on any topic — with live WebSocket thought streaming and per-report cost metrics.
LangChainWebSocketspdf-libFastAPI+1 more
Self-Correctingwip
Recursive Portfolio Chatbot
A self-correcting RAG agent that answers questions about my background using a LangGraph correction loop — with a 'See Thoughts' toggle showing internal reasoning.
LangGraphVercel AI SDKPostgrespgvector+1 more
Multi-Agentwip
Software Architect Crew
A 3-agent crew that takes a product prompt to reviewed code — PM, Coder, and Reviewer — with a live Mermaid.js flowchart showing which agent holds the token.
LangGraphCrewAIMermaid.jsOpenAI+1 more