Qwiksift

What it does

Qwiksift is currently in its MVP stage, built with Python (Django DRF) and React JS. The platform serves as an advanced ATS, enabling admins to post jobs through the Django admin interface. We leverage the Gemini API (gemini-1.5-pro-latest) to analyze job posts, extracting key details like job title, employer, skills, and qualifications. The Gemini API assigns weights to these factors, which are combined with custom algorithmic weights to help filter and rank resumes effectively.

When candidates apply, their resumes are parsed using Gemini LLM, and the extracted data is stored in a PostgreSQL database. For instance, if 100 candidates apply for a job, Qwiksift filters and ranks the top 20% based on the combined weights. Admins can then refine these candidates further by applying custom preferences, such as specific skills or location (e.g., Python, LLMs, NLP, and New York, USA).

The Gemini LLM, along with our custom algorithm, is then used to score candidates against these preferences, enhancing accuracy. Elasticsearch supports advanced searching, and the system re-ranks the candidates based on the new scores. Admins can apply up to 7 different filters to score and rank resumes, with the option to download the filtered results for further review.