Revolutionizing Recruitment with Large Language Models: A Multimodal AI Framework Integrating Video, Social Media, and Traditional Screening for Efficient Talent Acquisition

The classic problem with recruitment is the need to analyze multiple resumes in a short period while ensuring the best quality of hires. In the wake of Covid and the advent of generative AI, we have seen a modern approach to recruitment, digitizing the process to gather more suitable candidates for the institution. The process of recruiting candidates, especially for those in technical roles has lately gained a high rise. There has been a very rapid change to bring about non-contact based recruitment, helping institutions survey a wider range of people without limiting to geographical constraints. Newer recent changes seen since covid is the usage of online video chat platforms to conduct live one-on-one interviews. Our paper takes this approach of using video as a form of recruitment for screening candidates, we also delve into utilizing social recruitment also as a form of pre-screening using LinkedIn and GIT for technical roles. This paper also helps view pre-screening from traditional resumes and newer introduced methods by leveraging LLMs to help reduce human intervention until the final interviews. This approach helps combine the various diverse components into providing a solution to these recruitment challenges. The paper aims to help recruiters in selecting candidates more competently while improving time management, precision, objectivity, as well as reliability during the hiring processes.