Inspiration
We are all recent graduates or upcoming graduates with Computer Science degrees. We have seen how tough the job market can be and wanted to create something that would help recruiters make better and more informed decisions
What it does
This program allows recruiters to chat using Gemini to jobs and get insights into applicants
How we built it
We utilize VertexAI, Gemini, Pinecone, ReactJS and FastAPI to create a seamless experience for applicants and recruiters.
Challenges we ran into
We ran into a few challenges when creating our Retrieval Augmented Generation workflow. Namely, creating embeddings while also storing the PDF within MongoDB. In hindsight, it may have been easier to simply store the PDF embeddings rather than storing both, but we wanted the Recruiter to have the option of downloading the PDF just like any other recruiting platform would allow. We also ran into issues with system prompting and designing our VertexAI x Gemini workflow to work consistently across queries
Accomplishments that we're proud of
We created, what we believe, to be the first Recruiting forward RAG application built with Gemini
What we learned
We learned a lot, especially about working with LLMs and how RAG works to enable reduced hallucinations for businesses and consumers
What's next for Recruit Wise
We hope to be able to thoroughly test it further and potentially either open source it or make it a program that businesses can utilize
Log in or sign up for Devpost to join the conversation.