How Nova Talent Automated Their Candidate Review Process

Jun 3, 2023

Kevin Bartley

Customer Success at Stack AI

Nova Talent, the Global Top Talent Network, partnered with Stack AI to automate their inbound application review process.

They reduced costs by ~60% and increased inbound applications by 30%.

Opportunity

Nova Talent is a merit-based access network where the top 3% of talent connect, develop, and accelerate their careers. Over 500 top professionals apply monthly for membership.

Reviewing applications is crucial for Nova, as its value proposition relies on this process. Each application includes several video recordings of the applicant answering specific questions, a resume, and short responses to additional questions.

A Nova recruiting manager spends an average of 2 hours compiling and evaluating each application, making the process time-consuming and resource-intensive. Applicants typically receive a response after 2-3 weeks, a timeframe Nova is continually striving to reduce.

Solution

Nova Talent developed a sophisticated AI system in Stack AI capable of transcribing videos,

analyzing content against evaluation criteria, and providing an evaluation summary.

The AI system assessed the videos, resumes, and short answers to deliver a final recommendation.

After experimenting with different models, Nova Talent adopted a combination of Deepgram models for transcriptions and OpenAI GPT-4o for evaluations. With Stack AI, there's no need to set up accounts with these model providers; we handle it for you, and API calls are covered by our subscription.

The system was automatically triggered upon receipt of a new application, delivering results to the recruiting coordinator's inbox in less than 5 minutes.

This email contained the recommendation, evaluation report, and links to the videos, original resume, and answers for final audit.