Strong Results for LearningBranch’s A.I. Soft Skills Engine at Contact Center.
We’re currently carrying out a four-month testing phase at a large contact center in the Philippines, and results so far indicate that our A.I. Engine can effectively replace manual processes throughout the recruitment and training processes.
LearningBranch’s proprietary A.I.-powered soft skills assessment engine is in development, with the purpose of insightfully evaluating a candidate’s spoken and written communication skills according to our signature framework of crucial soft skills criteria. The technology uses A.I., natural language processing and our framework to score a candidate’s customer service, sales, and teamwork skills as well as spoken and written proficiency.
During the pilot phase currently underway in the Philippines – a key global market for contact center employment – candidates write or record their answers to a series of simulated customer inquiries. Responses are then scored by both our A.I. Engine and by external human evaluators. Results show that the A.I. Engine was found to be accurate over 75% of the time compared to manual scoring. The other 25% of the time, the A.I. Engine was within a 15-20% tolerance.
Manual Subjectivity vs A.I. Consistency
During our work with clients, we’ve observed a lot of subjectivity when it comes to in-house manual scoring of test responses, especially a bias to pass candidates. Our A.I. Engine brings consistency and standardization to the hiring process, as well as a great deal of speed and efficiency.
In addition, LearningBranch’s automated proficiency rating is correlated with the client’s inhouse training performance.
The tool can currently evaluate 4 specific foundational soft skills, including:
- Attention to detail
- Expressing empathy
- Refusing requests
- Using positive language
We have researched and defined 26 other soft skills that our engine could eventually assess to add further depth.
We have just started another pilot project with a global customer support organization that will allow us to further calibrate our A.I. Engine against associate performance on a global scale.
Any questions about our A.I. Engine? Let us know!