project
JBCN Admissions
Schools get thousands of applications but only a percentage of seats. I made agents that fit the pipeline.
Schools get thousands of applications every year but, they have only a percentage of available seats. Today these applications are assessed largely with humans. I made enterprise grade Agents that would fit right into the pipeline and use decision making context to rank leads.
This project gave agents tools for making sense of data and grading it consistently, used RAG for data lookups, telemetry to measure agent metrics such performance and cost, all tied together with Google's ADK. A task that would spread across weeks and multiple humans was now standardised with agentic AI.
My first project as an end-to-end application developer. After spending 7 years being a product manager in technical roles, I wanted to make something that was challenging 'cause only then would it be fun.