ETFProfiler

ETF Profiler is an intelligent, LangGraph-powered assistant that helps match investors with the most suitable ETFs based on their personal financial goals and preferences.

Business Case: https://sdw.solutions/case-studies/etf-profiler-webapp

It works by combining:

• 🧠 Conversational profiling: The agent asks users a series of questions to understand their investment style.
• 📄 PDF document ingestion: ETF factsheets and documents are parsed and embedded into a vector store.
• 🧭 LLM-powered matching: GPT-4 interprets and scores each ETF based on how well it fits the user's needs.
• 🔍 Semantic retrieval: Uses ChromaDB to surface the most relevant ETF documents before scoring.
• 🛠️ Built with LangGraph: Every step is modeled as a LangGraph node for full transparency, debuggability, and customization.

Source code: https://github.com/nikpalumbo/etf-profiler