How We Reduce Recruiting Costs Using Databricks Foundation Models

An organization's recruiting pipeline is time-intensive, requiring the time and efforts of multiple individuals and groups, including recruiters, solution architects, engineers, developers, and managers. Using Databricks, we built a genAI application for use by our non-technical recruiters to evaluate the technical capabilities of a candidate based on a given resume. We apply a multi-agent approach, leveraging the Databricks Foundation Model API to assess a resume's alignment and recommend if the candidate should move forward in the recruiting pipeline. Resumes and job descriptions are ingested into Google Cloud Storage. Inference results from Databricks Foundation Models are stored in Google Firestore. We tie it all together with a containerized Flask application served via Google Cloud Run. We'll demonstrate the application with our current use case and share the learnings with implementing and leveraging Databricks Foundation Models in a production environment.
Talk By: Gary Nakanelua, Managing Director of Technology, Blueprint
Here's more to explore:
LLM Compact Guide: dbricks.co/43W...
Big Book of MLOps: dbricks.co/3r0...
Connect with us: Website: databricks.com
Twitter: / databricks
LinkedIn: / data…
Instagram: / databricksinc
Facebook: / databricksinc