Can technology identify China's top graduates?
Has the humble CV finally met its match?
Most people will have drawn up a curriculum vitae, or resume, at some point in their careers - that all important "path of life" document listing your job history and academic achievements, occasionally peppered with, come on admit it, a tiny bit of embellishment.
But tired of drowning under paperwork from thousands of candidates, and often unable to tell truths from half-truths, French cosmetics company L'Oreal has had enough.
It has chosen the world's biggest jobs market - China - to utter two words that would be music to the ears of beleaguered recruitment executives everywhere: "Goodbye CV".
This year, the 33,000 applicants for the 70 places on the company's Chinese graduate recruitment scheme have been asked to save themselves the paper, the printer ink and the pain.
Instead, they were asked to answer three simple questions via their smartphones.
China produces seven million job-hungry graduates every year.
The volume of interest means big companies must filter out applicants based on some pretty crude measures. One of the most commonly used is to look at the university that the applicant attended.
"In previous years we've really relied on CV screening as the first step, which is pretty common for a lot of companies," said Jacob Bonk, recruitment director for L'Oreal China. "When in fact what we're really looking for in students is raw talent."
"You want people that are a better fit to your company culture and the competencies that you look for, and they're not necessarily going to have that just because they went to a particular university."
The idea that top companies recruit from top universities is not unique to China. It remains a reason students are so keen to get into such institutions. And in many professions it still holds true.
But there are plenty of experts who question how much high academic achievement is a predictor for success in the workplace.
Lazslo Bock, the man in charge of hiring for Google, suggests in this New York Times interview that super-successful students may lack a fundamental attribute; the ability to learn from failure.
And in many societies - including China because of schooling restrictions placed on children of internal migrant workers - the education system is stacked against students who are both bright and poor.
L'Oreal has long been aware of a big problem with its CV-based recruitment campaign in China. It is just not delivering enough of the kind of employees they need.
Scientific approaches to recruitment are nothing new. Aptitude tests and verbal reasoning models can be used to try to identify key attributes sought by an employer.
More recently there have been developments in the use of "big data", analysing the information stored about large numbers of employees in an attempt to help companies find recruits that are most likely, for example, to be loyal.
But the Shanghai based start-up company that is now helping L'Oreal in China believes it has something much more exciting to offer.
"We have developed algorithms that can take the words that people use and derive context from them," said Robin Young, the founder of Seedlink Tech.
"Language can be a very good predictor, for instance, of how intelligent someone is, how experienced they are, how much knowledge they have."
Rather than complementing the existing CV-based selection system, Mr Young claims he can replace it, and L'Oreal China seems to be convinced.
Here's how it works: students use their mobile phones to access L'Oreal's website which prompts them to answer three open-ended questions.
For example: "If you had one month and a 25,000RMB budget ($4,000; £2,570) to tackle any project your little heart desired, what would you do?"
The answers, which have to be at least 75 words long, are automatically fed into Seedlink's database and the software gets to work.
It analyses the language used and compares each candidate's answers with the many thousands of others.
Then, supposedly calibrated to mine for the specific personality traits that L'Oreal is looking for, it produces a ranking with, in theory, the person most suited for a career at L'Oreal at the top.
Out of 33,000 entries, 22-year-old Laurel Sun ranked sixth overall for her answers, including this one to the question posed above:
Laurel Sun's application answer
I want to set up an online shop that sells desserts to college students. The shop provides a door-to-door delivery service, but students only need to order on WeChat. The details are as follows:
- 1. Carry out an online survey: investigate students' favourite flavours and the prices they're willing to pay (RMB1000/7 days).
- 2. Accordingly, make a budget plan to determine the products, purchase the equipment and the raw ingredients and make the desserts myself (RMB15000/7 days).
- 3. Set up the e-commerce shop and have a one-week promotion, providing free samples or giving rewards to those who press the 'like' button on WeChat (RMB8000/14 days).
- 4. Launch the shop.
The computer loved it.
"Based on her CV, I don't know that she would have made it through the door," said L'Oreal's Jacob Bonk. "But when you meet her, you can tell right away that she is really a L'Oreal profile."
Of course, the company doesn't need to rely completely on the rankings produced by the computer.
This year it used them to select 500 candidates for a Skype interview from whom a smaller group of 200 were invited for face-to-face assessment, at which the CV was still used as a reference tool.
But by dropping the reliance on it for the initial screening stage Mr Bonk claims around a third of successful candidates this year are from universities they wouldn't previously have considered.
"We really saw a diversity in the student profile that we've never had in the past," he says.
"We're meeting students from universities that we've never looked at before and meeting profiles that are totally different in terms of style and cultural fit."
These are early days and L'Oreal won't be able to truly judge the success of the new scheme until it can compare the performance and, in particular, the turnover rate of its new recruits with those chosen previously.
But who knows, perhaps it will herald the day in which we're all chosen by computer.
"CVs just aren't doing their job to predict how well someone's going to perform in a career," Robin Young from Seedlink Tech argues.
"Computers have no emotions, they have no moods, so you can programme them to act like an ideal human and judge these answers without any sort of colour or any sort of bias."