video1.0<iframe src="https://www.loom.com/embed/4e4c1b3141bf4d008658999887775c8d" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/4e4c1b3141bf4d008658999887775c8d-9d177dd124197a2d.gif258.921In this video, I present my candidate search direction project, which I believe is more complex than standard job searching. I built a four-stage architecture that processes raw unstructured data and retrieves the top 20 candidate profiles based on a refined query. The system uses advanced techniques like semantic routing and progressive relaxed sieve to handle negations and optimize search results. I demonstrate the system's capabilities using a sample query and highlight how it effectively filters candidates. I encourage you to explore the results and provide any feedback.