woman on phone SH004 v2

Enhance Your Candidate Experience by Adding AI to Assessments

How SHL, powered by Aspiring Minds, uses Artificial Intelligence to identify candidates with strong language skills, find the best coders, and simulate on-the-job experiences.

Assessments have never been more important. Businesses can no longer afford to eat the cost associated with “gut feeling” hiring. In addition, talent is hard to come by with global unemployment rates at an all-time low. So, when you hire an employee it is more and more important to assess for future potential and not just for the open requisition. Skills tests are among the suite of highly sought after assessments these days. Assessing for motivation, personality, and cognitive ability is incredibly important, but adding a skills assessment will guide you towards someone that has the full potential you are looking for.

But have you ever taken a skill test for a job? Often it is a list of time-consuming multiple-choice questions that are not all that exciting to answer…

Artificial Intelligence (AI) is changing the experience of skills assessments and changing it fast. With AI-scored assessments, candidates have an opportunity to express their creativity and showcase how they work through functions like writing an essay or an email, or by talking to a customer or programming a website. They get a real experience of what they will do on the job– replacing the feeling of test-taking all together. And, guess what – these assessments or shall we say, experiences, evaluated using AI, provide a more accurate assessment than the conventional multiple-choice questionnaires (MCQs). A win-win for businesses and candidates!

Here are a few examples of what AI has done and is doing to assessments, a vision inspired by our eight years of work:

Measuring for Language and Writing

There are some things that MCQs just cannot evaluate. For example, it is difficult to adequately measure the fluency or pronunciation of a candidate’s language abilities until you hear them speak the language. Similarly, it is difficult to know a candidate’s writing capabilities until you have them submit a writing sample.

Through our AI technology, we give candidates an opportunity to demonstrate their spoken language skills by talking to a test or writing an email in response to another. Using speech recognition, natural language processing, and machine learning, we generate scores on various job-related competencies such as grammar, content relevance, email etiquette, and speaking skills. The machine-generated scores closely mimic expert grades.

Our products SVAR and WriteX, as presented in numerous publications1 at top AI conferences, are used across the globe and have helped companies reduce recruiting processing time from 2 weeks to as little as one day2. Strong language skills lead to higher income for candidates and are required in a large number of jobs across the globe. This is one of many examples where AI makes the employee lifecycle more efficient and a better experience for everyone.

Through our AI technology, we give candidates an opportunity to demonstrate their spoken language skills by talking to a test.

Finding the Ace Coders

Coding simulations have been available for a decade. However, there is a major issue because talented coders often receive a zero score due to an immaterial error, leaving them out of the candidate pool. This problem exists because coding simulations are not able to give a partial grade – but just a 0 or 1 score - making the whole coding simulation behave like an MCQ – you spent 30 minutes writing it, but either you got it right or a zero.

Over the last five years, we have created the most sophisticated engine for grading code. Backed by three KDD and IAAA papers3, our tool can grade partially correct codes and even those that do not compile. The grades mimic how a human interviewer would grade the code, looking at its logic and not necessarily swayed by syntactical or inadvertent errors. The tool improves hiring throughout by 30-60% in different scenarios.

Furthermore, we do a deep analysis of the code to provide rich insights like programming practices—whether the candidate writes code that could be vulnerable, hard to understand, and non-modular. It further analyzes whether the code is time-efficient or it takes forever to execute. Such insights are a great tool for engineering managers to hire highly productive engineers and software designers. Our coding assessment, Automata, is used by the world’s largest eCommerce, internet and services organizations across the US, China, and India.

Over the last five years, we have created the most sophisticated engine for grading code.

Engaging Call Center Simulations

Hiring chat agents in a call center is another example of how AI is advancing the hiring process. Here, the applicant experiences chatting with a simulated customer, who say, has lost his credit card. The customer talks like a real customer would do, asks and responds to questions in natural language, and even nudge one if they aren’t responding. The applicant uses a simulated information system, follow the standard process, search information and solve the customer’s problem while being polite and customer focused.

In this simulation, they are experiencing actual work, even interacting with a customer, and are not filling out bubbles on a test. The recruiter gets deep insight into how the candidate will do the job: adhere to process, follow chat ettiquites, and resolve queries in time.

The Future Is Exciting

AI is surely a win-win for customers and candidates. So much progress has been made to create engaging and immersive experiences for employers & candidates through the use of AI. And we are not done innovating yet! We desire to continue building products that provide value to our customers as well we their employees and candidates. Our dream is to help business utilize people insight to truly utilize their greatest asset – people!

Contact us today to learn more about how we can help you attract, develop, and promote the best talent.



1 Automatic Spontaneous Speech Grading: A Novel Feature Derivation Technique using the Crowd - Vinay Shashidhar, Nishant Pandey, Varun Aggarwal - ACL 2015

2 90% reduction in candidate processing time by using AI-powered spoken English assessment [https://www.aspiringminds.com/case_studies/90-reduction-in-candidate-processing-time-by-using-ai-powered-spoken-english-assessment/]

3 Aspiring Minds’ publications [http://research.aspiringminds.com/publications/]


headshot varun aggarwal


Varun Aggarwal

Varun Aggarwal is a researcher, entrepreneur, and author. In 2008, he co-founded Aspiring Minds and now heads Research. His work has led to the world’s first ML-based assessment of coding skills and the world’s first automated motor skills assessment. Varun has published more than 30 research papers and his work has been covered in The Economist, MIT Technology Review, Wall Street Journal, HT Mint, Economic Times and IEEE Spectrum. He started ML India (www.ml-india.org) to build India’s AI ecosystem and organized the world’s first data science camp for school children. Varun was awarded the HUMIES award in 2006 for developing algorithms that mimic human intelligence. He enjoys writing poems and stories.

Explore SHL’s Wide Range of Solutions

With our platform of pre-configured talent acquisition and talent management solutions, maximize the potential of your company’s greatest asset—your people.

See Our Solutions