Watson alum builds AI for HR

​If you were to choose 10 of your favorite photos from your social media accounts, what would they say about you?

Maybe they’d show you are a city-dwelling runner, devoted dog-owner and foodie, who had a perfect-hair day on April 18, 2015.

Ganesh Iyer, MS ’05, says that what he would see has nothing to do with your hobbies, your hound or your hair. What he — or, more precisely his startup, Dotin.us — would see are personality traits that can predict how well you’ll fit in at the company where you’ve applied for a job.

A mismatch between employee and employer can be a disaster for both. But a good fit need not rely on just interviews, skill sets and luck.

Iyer calls Dotin a “personality DNA company” that uses artificial intelligence (AI) to improve matches between employers and employees, and marketers and consumers. And it starts with 10 photos.

In broad strokes, here’s how it works:

A user uploads 10 favorite photos to Dotin.us, which uses an algorithm to read just the colors. The photos are essentially stripped of facial expression, race, gender, ethnicity and contour, which eliminates the potential for bias.

A human resources department using Dotin does not know what you look like, nor can you try to influence a decision by curating an image or attitude.

The colors (further sorted into common combinations and contrasts) are mapped to personality traits, which can then be matched to the learning and working styles necessary to succeed at a particular job.

“If we do that right, we help the employees grow their careers and we help the company grow their talent,” Iyer says.

Tests have built-in bias

Companies have long used personality and skills tests to assess potential hires. Not only are the tests long, they’re flawed, Iyer says.

For job-seekers, the moment you start answering an interview question, your conscious mind kicks in and your answers might not reveal the core motivations of who you are and what you might bring to the table.

“You’d always give answers that would be conducive to the desired outcome. That’s how humans are wired,” he says.

For the employer, Iyer says, the question is, “How do you remove that bias? The only way is to never ask a single question and rely on subconscious motivations only.

“That’s one of the reasons we started looking for the subconscious influences that go into your decision-making process. And the most profound one is color.”

An oft-cited statistic is that people make up their minds within 90 seconds of their initial interactions with either people or products, and that 62 percent to 90 percent of the assessment is based on colors.

Marketers have used this knowledge for a century, Iyer says, but deriving personality type from colors is no longer just fodder for quizzes in women’s magazines, and that is because of AI and machine learning.

Relying on a team that includes data scientists, mathematicians and a psychologist, Dotin has, so far, identified 1,560 color combinations and mapped them to 150 personality attributes that are mapped to the talents, learning and working styles in business, as identified by Gardener’s theory of multiple intelligences and Holland’s theory of career choice. Dominant and compatibility attributes are included. Those personality attributes are linked to the following desirable business outcomes: compatible cultural fit for a job or a team, improved performance, increased productivity, increased retention and improved manager effectiveness.

The end result is that employers can be highly selective in their hiring search, and job-seekers can find positions that are well suited to their personalities. The process can be done for existing employees, too, giving companies better insight into the key attributes for specific jobs and helping employees identify mentors or training to help them in their careers.

Jayashubha Kunnummal, chief people officer for TVS Credits, one of the largest financial companies in India, has been working with Dotin since spring 2018. She was intrigued by the concept of identifying personalities and then using that to make decisions about hiring, promotion and training.

“We have tested his product to find out its validity … and found it is relevant to our organization,” she says. “We have been able to see a difference in personalities of high, medium and low performers at the front-line staff level.

“The identification of learning style for an individual and customization of training is very interesting and useful,” she says.

In the United States, Dotin is helping San Jose University students and employees find mentors. In Canada, Intrideo has integrated Dotin into its candidate tracking system for the HR and hospitalities sectors. (In its first study, Intrideo reports that it has helped A&W franchises reduce understaffing and increase employee retention.) In the United Kingdom, PwC has given Dotin a thumbs-up for being unbiased and ethical, and admitted it into its Scale | GovTech program (like a U.S. accelerator program) with a focus on transforming delivery of government services.

From engineer to entrepreneur

Iyer graduated from Binghamton in 2005 with a master’s in mechanical engineering. His first job was with a startup in the San Francisco Bay Area, where he used his microelectronics expertise to re-engineer parts to withstand atmospheric stresses when used in satellites.

“I was very happy that I didn’t have to talk to people. I was happy being in a lab, doing my work and publishing papers,” he says.

A few months into the job, he accompanied his boss to pitch a contract to Raytheon. The plan was this: His boss would lead the pitch, Iyer would provide some slides and technical explanation, and the day would end with drinks and, hopefully, a contract.

Except the boss was called away for an emergency in the middle of the presentation, leaving Iyer to run the show. And take them to dinner, too, his boss added.

“I literally lost words,” Iyer says.

“In the evening, I had to take these 20 executives to a bar and start socializing, and it was next to impossible for me. I was so scared and nervous. That was the first time I realized how introverted I was.

“I also wondered if there was a tool out there that could help me be more prepared to talk to people I don’t know,” he says.

That idea stuck with Iyer, who moved on to Cisco Systems, where he held a variety of jobs. His last position was with the sales team.

“As Google, Facebook and LinkedIn became more pervasive, I used to spend 10 to 15 minutes doing research to understand who clients were so I could start a conversation,” he says.

Then one day, a manager missed a meeting, and what happened to Iyer in his first job happened to a colleague.

“He did a terrific job selling the technology, but he couldn’t connect with any person in the room. And we lost the deal.

“That was my ‘aha’ moment. This was not just my pain point, it’s a pain point for many people.”

So, he began working on an app that would understand the user’s personality and make recommendations on how to talk to people nearby. He worked with psychological experts and data scientists.

What became clear to them, he says, is that people shape themselves depending on their needs, and the only way to remove that “mask” is to identify their subconscious motivations.

Discovering the work of Shigenobu Kobayashi, Japan’s leading color psychologist, and subsequent research linking color and personality, Iyer and his team realized they were onto something far bigger than an app.

“We started getting a lot of inquiries from enterprises: If you are able to do this, can you help us understand more about our employees? Can you create some analytics?

“Feedback from customers is what has driven us to take this tack,” he says.

Dotin.us debuted in 2014.

Dotin takes off

“The rise of Dotin.us begins now! Send us your love if virtual or cheer us live!” Iyer posted on Facebook in July from Hong Kong.

In August he was in Bangalore, India; and San Francisco, Calif. In September, Las Vegas and Northwestern University in Illinois; and in October, Tokyo, Japan; and Ponte Vedra Beach, Fl. Each location was a conference or expo.

In early March, Dotin was named “most innovative startup” among eight finalists at the Plug and Play Japan demo day. Plug and Play holds events around the world, bringing together startups and large corporations. Demo day judges came from companies including Nissan, Honda, Aflac, Panasonic and NEC.

Iyer has participated in accelerator programs, secured investments from NetOne Systems and serial entrepreneur Amar Chokhawala (founding member of Google Adsense) and in November received his first patent for Dotin.

Each event helps Iyer build his customer base and collect more data to work with. There is no downtime for a startup, and barely time for personal reflection. But did he ever, as an engineer, envision life as an entrepreneur?

“Oh man, I am still surprised where I am,” he says, laughing.