Large data recruitment, so I was selected by the algorithm

Jede Domingues, a 26-year-old who has never been to college, was judged by the gold-plated company's algorithm as a programmer, and then recruited by the new company.

Last summer, 26-year-old Jede Domingues (Jade domingues) received a sudden e-mail that a San Francisco start-up company asked him to interview programmers. Dominguez, who lived in a rented room in Pasadena, California, on credit cards, is learning to program. Dominguez in high school, and did not want to go to college. But somewhere in the clouds there was a man who thought that Dominguez might be a genius, and that it was a rough stone.

"The traditional indicators used in hiring may be wrong, big mistake," said the chief scientist of the gold-plated company, Vivian Ann Ming.

Gold-plated company founder Luca Bommassat and Hillrod De want to automate the process of discovering good programmers.

That person is Luca Bommassat (Luca Bonmassar), who discovered Dominguez through a technology that will raise important questions about how companies are recruited and whether good people are missing out on the process. The new idea has made people look away from the traditional talent index, for example, recruiters are generally interested in MIT degrees, Google company experience, colleagues or friends recommended, and so on, while devoting more attention to some simple concepts: how is this person's performance? What can this man do? Can you quantify it?

The technology is the product of a gold-plated company (Gild), a start-up that is less than two years old, and Bommassat is one of the co-founder. A small group of new companies such as gold-plated companies is running the goal of automating the discovery of talented programmers-a group of huge market demand. Their work falls into the category of large data that uses computers to collect and analyze a wide variety of information, perform numerous tasks, whether it is recommending books, placing targeted ads on the site, or predicting therapeutic effects or stock prices.

Let algorithms help you find people who are not found

Recently, more and more scholars and entrepreneurs have applied large data to human resource management and talent search, creating a new field called Labor Science. What gold-plated companies are doing is to see if these technologies can also be used to predict the performance of a programmer at work. Gold-plated companies are scouring the internet for clues: Did he or she write code that was well received by other programmers? Is the code reused? How does this programmer communicate ideas? How does he or she get along with people on social networking sites?

The gold-plated company's approach to a large extent is still only in the initial stage, the effectiveness of what remains to be tested. There is a lot of constructive controversy about the idea of using big data to recruit, as well as excitement, especially in industries where it is hard to find talented people.

Gold-plated companies are expected to earn about 2 million to 3 million dollars this year, and the company has raised around $10 million trillion in financing, including a large sum of money from the early investors of LinkedIn and Cavan Mark Kvamme, a venture capitalist. There are big-name customers who test or use the technology of gold-plated companies, including Facebook, Amazon, Wal-Mart, Google and Twitter.

Companies use the technology of gold-plated companies to unearth new candidates, while also assessing the candidates they already consider hiring. Gold-plated companies have used the technology themselves-they desperately need competent programmers, and can pay out more than the big technology companies, so they found Jed. The algorithm determines that Jed's programming score is the highest in Southern California, and is 100 points to almost no one's enemies.

Who's Jed? Can he help the gold-plated company? His story tells us about today's recruiting system and meritocracy.

Ability to decide everything?

Most people in Silicon Valley agree that certain ideas, such as development, efficiency and speed, are good. Technology can solve most things. Change is unavoidable; unrest is nothing to be afraid of. Perhaps one of the most believed in Silicon Valley is the ability to decide everything.

But Vivienne Ming says she thinks Silicon Valley is not as meritocratic as people think. Vivian, who was the chief scientist of the gold-plated company since the end of 2012, believes it is common for talented people to be overlooked, wrongly sentenced or missed. She thought that partly because she had experienced it herself.

Dr. Ming is a male, named Evan Campbell Smith (Evan Campbell Smith). He was a good student and an excellent athlete, keeping track of high school athletics triple and pole vault far away. But he always felt that he had some sort of detachment from his body. After graduating from high school, Evan went through a full-blown identity crisis. His university ended in failure, in exchange for a change between jobs, thought of suicide, fell to the so-called life's lowest. But he didn't get stuck there, but he bounced again. At the age of 27, he returned to school with an undergraduate degree in cognitive neuroscience from the University of California, San Diego, and subsequently received a Ph. D. in psychology and computational neuroscience at Carnegie Mellon University.

During his postdoctoral post at Stanford University, he began a gender shift and became a complete doctor of Vivian in 2008.

As a woman, Dr. Ming began to notice the difference in how people treated her. Some are seemingly innocuous things, like men opening doors for her. There were also things that bothered her, such as the number of students asking her math questions less than when she was a man, and the number of male and female associates who asked her to socialize (like a baseball game) less often.

Prejudices are often manifested in forms that are not in the consciousness of the people. Dr Ming cites a finding by Yale researchers that faculty members of research-oriented universities describe female job seekers who are candidates for a manager position as less capable than men with the same qualifications. Another study, published by the National Bureau of Economic Research, found that job seekers with the name "sounds like Black" were more likely to receive an employer's call back than a "white-sounding" name on their resume.

Basically everyone agrees that the pronunciation of sex, looks or names should not affect hiring decisions. But Dr. Ming took the concept of meritocracy further. She suggested that the accepted criteria for screening talent, such as where to go to college, where they had worked before, would also miss out on talent and eventually become an employer's loss.

"The traditional indicators used in hiring may be wrong, big mistake," she said.

For what she called "so many wasted talent," Dr Ming's response was to make machines that would eliminate human biases as much as possible. This is not to say that the traditional sense of entitlement should be ignored, but rather to balance the measures she considers more complex. Overall, the gold-plated company's algorithmic analysis of a person to deal with 300 or so major variables: frequently visited websites; Describe the type of language used in various technologies, positive or negative, the skills on LinkedIn, how long you've been involved, and--yes--where you learn, what you learn, What was the ranking of the school in the American News and World Report?

Looking for gold in the data

Gold-plated companies are not the only ones that have been panning for information. Another San Francisco start-up, Talentbin, also searches for talented programmers on the internet, according to the company's Web site, Talentbin, a network of talented programmers gathered on the web, collecting "data waste" to create a potential recruiting list for employers. Another competitor is "Remarkablehire", a company that assesses individual abilities by looking at how his or her online results are scored.

And Entelo, the company is trying to find people who might be looking for a job, even before they start looking for a job. According to its website, Enetlo uses more than 70 variables to find signs of career change, such as the way a person displays himself on a social networking site. "We deal with data and save you Energy," the website wrote. ”

Susan Aitlin, an analyst with data and analysis at Altimeter Group, a UK data analyst, said it was "definitely worth a try" to apply big data in recruiting. But she raises questions about whether the algorithm improves what employers are already doing, such as collecting resumes or referrals, using metrics that are traditionally relevant to success, and so on.

"The lack of actual results," she said, "is the probability that the reality is not convincing me." ”

Sean Gourley, co-founder and chief technology officer of Quid, a big data company, said that screening data could provide information for hiring, but only if it was used to understand information that was not disclosed by the data. "Big data has big data biases," he says. "You measure what you can measure," and "you underestimate things that cannot be measured, such as intuition and charisma."

Gourlay added: "When you exclude people from complex decisions, you can optimize the algorithm into a god, but what is the price?" ”

Dr. Ming did not say he wanted to eliminate judgment, but she did think it was up to the computer to take the job, just like an automatic talent absorption and a sieve detector. Gold-plated companies have amassed a database of 7 million programmer data and ranked them according to the so-called gold-plated company score, which says the gold-plated company score is a measure of what a person can do. The ultimate goal, Dr Ming hopes, is to broaden the algorithm so that it can be used to search for and evaluate various types of labor, such as web designers, financial analysts, and even retail sales staff.

"We have dug up a gold mine within ourselves," Dr. Ming said, "and we found a computer-crunching kid in Los Angeles." ”

She was talking about Jed.

A person found by the algorithm

Dominguez grew up in Los Angeles and ranked third among five children. Mom is a housewife, dad is a telecommunications equipment installer, is an education-oriented blue-collar.

Jed's growth trajectory was rebellious. High school read half, has been a straight-a-a-student small Dominguez began to think, go to school in the end is to achieve standards or to real learning. "Value proposition (proposition) goes to school to get a good job," he told me. "But calm down, don't you want to go to school to learn?" Jed's grades dropped sharply, and he told me that he graduated from Alhambra High School with an average score of less than 3.0.

Not only did he not want to go to college, but he also wanted to prove that he could be super successful without going to college. He read a lot of entrepreneurial books, opened a custom design in the T-shirt printing company, first at home, and later moved to a rental of nearly 100 square meters in the warehouse. He thought he needed a website, so he taught himself to program.

"I was trying to prove myself with my own strengths," he said. He admits he may have done a little too much. "To prove that people are wrong and do things is not very mature performance." ”

He had a tattoo on his arm, "Believe" (believe), which was written by the flower body. Now he thinks it's funny, but still thinks he can do what he wants to do. When it comes to computer language, he says: "The best thing about code is that it is largely driven by ability." Don't look at what you've learned, just what you've learned. ”

When gold-plated companies started hiring, it assumed that the people of San Francisco and Silicon Valley had been cleared. So the company let its algorithm run through the Southern California information, drawing a series of programmers. The first one is Dominguez, he GitHub a very solid reputation on the site, GitHub site is where software developers gather, where they share code, exchange ideas, and build fame. The gold-plated company combed the information on GitHub and BitBucket, Google code and other few websites, looking for smart people to do it.

Dominguez's achievements. He wrote for a jekyll-bootstrap code, the site will be used to build a function, has been 1267 of other Web site developers reuse, leaving a very deep impression. His language and habits show enthusiasm for product development and a passion for a variety of programming tools, such as rails and JavaScript, that are useful to gold-plated companies. His comments on his blog and on Twitter suggest he is adamant that the gold-plated company wants its start-up members to have it.

The gold-plated company recruiter sent him ane-mail to ask him to come to San Francisco for an interview. The two founders of the company met a charismatic, confident young man---------------------------------------------------------

Dominguez wore a bright green hoodie to interview. He asked sharp questions, such as how the company rated the engineers without their knowledge, and whether they feared it would be considered a violation of privacy. (Gold-plated companies do not think so, Dominguez do not believe.) Gold-plated companies say it uses public information. ）

They also asked him some very mild questions, such as whether they could work in an organized environment. He said he could. The company immediately signed the Dominguez, and he received a salary of about 115,000 dollars of work.

"He's a typical, energetic person, but for whatever reason, he didn't get mobilized in high school and didn't see the value of college," Desai said.

Mr. Desai went to college and was one of the most highly respected institutions of the Massachusetts Institute of Technology. He was there to learn how to cope with stress, and to work with talented people, he said, and he had a time to sigh. While it's important to learn at school, he says, "That's not the whole story." Although he has a degree in computer science, he says with certainty: "I'm a lousy developer." ”

Large data recruitment, how reliable?

David Lewin, a professor of human resources management at the University of California, Los Angeles, says asking what a person can do is an important question, and it's equally important to ask him if he can do it with others. Dr. Levin says the surest way to predict how a person behaves in an organization is to make referrals to people who have already worked there. Current employees are aware of the company's culture and will be implicated in their own reputation and working environment. A recent study by the Yale School of Management, which uses large data, has refined the existing understanding that staff referrals are a good way to recruit good employees, but this approach is often effective when the recommender itself is highly productive.

In Dr Levin's view, he doubted that the algorithm would completely replace the proper recommendation of a reliable employee.

The customer of the gold-plated company has a company called Square, which operates a mobile payment system. Square, like many other High-tech companies, has been hiring, and Brian Bauer, the company's human resources director, has been in Silicon Valley for years, and according to him, the competition for good talent is as fierce as the dot.com boom.

"Stanford or working at Google is an excellent indicator," Bauer said. "It makes sense to be famous. "But there are a lot of options for these seed candidates, and they don't necessarily choose square. "We need to cast a net in a bigger pool," he said, "and that's what gold-plated companies are doing." ”

The technology of the gold-plated company found some candidates for square, but it hasn't been signed yet. Mr. Bauer said the gold-plated company's algorithm gives a generalization of programming scores, and that the requirements of square to fill vacancies are not so specific. "Gold-plated companies have a view of who has the ability to have it, but things are not that simple," he said, while saying that square was talking to the gold-plated company about optimizing the model.

For the moment, though it is of limited use, Bauer says the gold-plated company is doing things at the start of a big event. Today, young engineers are increasingly publishing their work online, participating in open source work, and providing more data that can be used to excavate raw stones. "It's all about finding people who haven't been discovered," he said.

Dominguez, who has worked in the gold-plated company for 8 months, has proved to be a talented programmer, Desai said. But he also said that Dominguez "sometimes struggle to work in a structured environment." When he sits in front of the computer, his colleagues try not to disturb him.

At the meeting, Dominguez will speak enthusiastically. He is happier than ever, he said, "As long as I can have a say in the process of building a system," or else he will be a system he can only follow. He has also complained a little about the company's expansion, which has expanded from 10 to 40 in the past 6 months, adding management and red tape.

"The truth is, my character is to do things in my own way, and eventually I have to start my own business," Dominguez said. However, he added immediately: "I appreciate and respect the opportunity the company gave me, and I know that they hired me because of my ability." I will always be grateful for that. ”

Dr. Ming said Dominguez this young man was a great discovery and an unknown. Of course, he is just an example that cannot be used to support or refute this new approach.

"He always wears that lone wolf," Dr. Ming said. "It's okay at the beginning, but it might be hard to say. ”

The algorithm is excellent at measuring what it can measure. It Dominguez the ability to deal with computers. It is still uncertain how he will use his talent to work with people in the long run.

Compiled from: "The New York Times", how the big Data are Playing recruiter for specialized brought

Article Picture: nytimes.com

(Responsible editor: The good of the Legacy)