AgentOctober 10, 2026
Product recommendation agent
What it does
Turns a shopping question into ranked product picks with a short explanation of why each fits.
Why I built it
To build an agent as an explicit graph, where each step can be tested and swapped on its own.
What goes in
A natural-language request, e.g. a product need and a budget.
What comes out
A ranked list of products with a one-paragraph explanation.
How it works (steps)
A LangGraph state graph with five nodes: analyse the intent, search a FAISS index of products, refine the results, explain the picks, format the response.