CareerGPT - Career graphs + AI agents to find your next move

Over the past semester, my team and I have been building Neo4j and knowledge graphs into CareerGPT, and it’s made me even more convinced that graph databases are underused in the agentic AI conversation. One reason is reliability. A lot of modern AI stacks rely on traditional RAG, which is largely based on fuzzy semantic matching. That can work well for discovery, but it often struggles when you need precision and grounded reasoning. With a graph database, you can query explicit relationships directly using Cypher, which creates a much stronger foundation for trustworthy outputs—especially when the goal is to reduce hallucinations and keep results tied to structured facts. What surprised me is that interest in CareerGPT showed up even before the graph database story was front and center. A couple weeks ago, I shared the product and saw 40+ sign-ups from Georgia Tech students. It was a validating moment, and it pushed me to look more closely at the onboarding responses to understand what was driving that interest. A consistent theme was that many students were looking for mentorship and clarity in the age of AI. That led to a broader realization: even strong technical programs don’t consistently surface what students are really trying to understand right now—which roles are rising, which are cooling down, and what practical paths make sense from a given starting point. That’s why I built career graphs into the platform. The goal is to make industry trends more visible and help users explore realistic career paths based on where they are today. This is presented to users in an easy-to-digest graph visualization. Since AI is affecting every industry, CareerGPT isn’t just for students anymore—I’ve removed the .edu gating so anyone can try it!