RiverGuide: AI-Powered Mentor Onboarding

RiverGuide is a mentorship platform matching University of Louisville students with field-specific mentors. In a separate engagement, I was brought back to solve a specific bottleneck: mentor onboarding was too slow and too manual.

Getting mentors onto the platform required them to manually fill out detailed profiles covering their expertise, background, availability, and mentoring focus areas. This created friction that slowed down the supply side of the marketplace. Mentors would start the process and drop off, or submit incomplete profiles that weren't useful for matching.

I designed and built an AI agent that extracted a user's professional details directly from their LinkedIn profile, then auto-generated a complete mentor profile they could review and edit before publishing.

The agent handled the heavy lifting: parsing career history, identifying areas of expertise, and structuring everything into the format RiverGuide needed for its matching system. Mentors went from a blank form to a fully populated profile in seconds, then just tweaked what needed adjusting.

The pipeline was built with FastAPI and Python on the backend, with a web frontend that gave mentors a clean editing experience on top of the AI-generated output.

50% faster mentor onboarding compared to the original manual flow

Reduced drop-off during profile creation by removing the blank-form friction

Built a reusable LinkedIn-to-profile extraction pipeline that could extend to other onboarding contexts