ÉCLAIR is an active Emory research project and an AI-assisted recruitment tool which automates the accurate and non-biased identification of suitable job candidates for open job description(s). ÉCLAIR was developed in recognition of the fact that matching applicants to open job positions is typically a manual, error prone, time-consuming effort, which relies on imprecise semantic searching, and/or unsophisticated screening methodology. ÉCLAIR benefits the Human Resource recruiter by providing an intuitive job suitability score which can be balanced with regulations thus allowing recruiters to retain control of the hiring process. Currently, ÉCLAIR is focused on the clinical research coordinator environment with the possibility of future expansion to other employment domains.

The ECLAIR Project is an Emory University-based collaboration between members of the Nell Hodgson School of Nursing, the Emory Department of Computer Science, the Department of Biostatistics and Bioinformatics in The Emory Rollins School of Public Heath, Emory Healthcare, and the Emory Office of Technology Transfer. We are currently funded by the Georgia Research Alliance using a Phase 1 grant.

The project website is available at http://steviep42.github.io/eclair

Project contacts are Elaine Fisher (elaine.fisher [at] emory.edu) and Steve Pittard (wsp [at] emory.edu)

‍

Fifty-two percent of talent acquisition leaders report that the hardest part of recruitment is identifying the ‘right candidates’ from a large and diverse applicant pool (https://ideal.com/ai-recruiting/). Our work has been motivated by this problem in conjunction with the idea that a well-prepared Georgia research workforce is necessary to implement regional health science research. A projected 10-14% increase in the number of Clinical Research Coordinators (CRCs) nationwide is expected by 2026. For Georgia, an 11% growth rate for CRCs (from 1,030 to 1,150) is anticipated. Job aggregate websites such as Monster, Indeed, and Glassdoor report position openings for CRCs between 1,221 and 9,714.

However, recruitment goals for CRCs can be frustrated by overly broad job descriptions which prevent the accurate matching of candidates to specific positions. At Emory University, on average, up to 300 Clinical Research Coordinator applicant resumes must be viewed to arrive at 30 possible candidates.

We believe our Natural Language Processing framework can facilitate a more efficient, accurate, and non-biased candidate identification of clinical research professionals when compared to approaches reliant upon manual review or simple semantic matching. CRC resumes can be interrogated using an advanced model that will normalize content into standard concepts for matching with stated position requirements. We will implement an ÉCLAIR prototype for initial use at Emory University with a focus on the clinical research coordinator domain.

‍

‍