Talk to the Agent — text, voice, or upload your resume
Quick Start — click any prompt to begin
How It Works
You start the conversation
Type or speak. Tell the agent about yourself — career goals, experience, or just say "I'm thinking about school."
RAG retrieves the right info
The agent searches a 77-chunk knowledge base covering 13 graduate programs, 5 personas, and 24 career goals.
Grounded, source-backed answers
Claude Haiku 4.5 synthesizes a response using only retrieved context — no fabrication, no hallucinated programs.
About the Project
This agent was built as the final project for MGMT 59000-ETB (Spring 2026) — Entrepreneurial Thinking and Behavior at Purdue's Mitch Daniels School of Business.
It uses retrieval-augmented generation over publicly available program information. The user-facing agent runs on Voiceflow; a parallel Python pipeline instrumented with LangSmith provides full observability and traceability.
All code, knowledge base chunks, system prompts, and trace evidence are open-source on GitHub.
View source on GitHub →