My family and I were recently hit with power outages that plunged our city into complete darkness for several nerve-wracking hours. It's an experience that taught us to take our dependence on a continuous supply of power less for granted.
According to their website, VIA "is using A.I. to help the world's power generation and distribution companies predict outages, improve the flow of power through their vast power networks, and otherwise help cities and countries keep the lights on for millions of people." VIA is working with some of the world's largest power transmission companies in the world, like TEPCO in Japan, which also holds a stake in the company.
Barely a year and a half after its founding, VIA is already earning accolades for its innovative technology. In September 2017, it accepted Startup Canada's 2017 Innovation Award for the Quebec region. The award "recognized VIA's leadership in using artificial intelligence to solve large-scale, high-stakes challenges across critical infrastructure, particularly in the energy industry."
Headquartered in Somerville, Massachusetts, VIA is backed by a group of angel investors which includes Boston-based NXT Ventures. In October 2017, it was one of just a dozen companies selected from more than 400 applicants to join Elemental Excelerator's 2018 Demonstration Track, an exclusive incubation program that will give them access to $1 million in funding to further develop their new blockchain technology for distributed A.I. learning in the energy industry.
As a serial entrepreneur who has launched, scaled, and sold several companies, Gounden has extensive experience recruiting and developing technical talent. I spoke to him recently about how he finds tech talent in what today has become one of the hottest areas: A.I. and machine learning.
You hire a lot of technical talent. What processes and techniques do you use to find people with the skills you're looking for?
Colin Gounden: Everyone talks about technology and IT, but it's people who create the technology and the IT. But how do you find, keep, and motivate people? That's the biggest imperative for us. Anyone can fool you in a 45-minute interview. But when we hire people, we actually give them tests. We'll ask them to take one of our company values from our website and write what that means to them.
For people applying to a data science position, we'll give them a case study and ask them to solve a problem. We give them just two hours to tackle it because we want to observe their approach to solving tough problems. There might be multiple ways to solve it. We want to hear how they think about the problem and how they frame it. We learn a lot this way.
AI is a skill that's in high demand. How do you find people with this background?
Colin Gounden: Since we already have a core group of experts that are highly experienced in A.I., we don't necessarily look for people who already have a lot of data science or AI experience. We actually hire people with PhDs in physics. What we want are people who are smart and motivated, and who are good at solving problems. You can teach programming and data science, but not innate things like motivation and problem-solving and creativity.
Of course, for technical roles, we don't hire art history majors. We want people who understand a circuit diagram and who are mathematically inclined. We don't want to have to teach them basic math or calculus.
How do you test for "softer" skills in candidates you recruit to see if they'll be a good fit with your company culture?
Colin Gounden: We test for two types of skills: The "hard skills"--does someone know how to setup an Amazon server, or write a program in Java. And then we test for the kinds of skills that others might call "soft skills," but what we refer to as the "harder skills"--which are way more difficult in our opinion.
These include things like being able to communicate effectively. When someone gives you a task, what's your ability to say "no." For example, from my previous experience running a company that had an office in India, it's often considered impolite to say "no" in that country. So they might say "yes, I can" even if they can't actually do a certain task they've been asked to perform. But this is not a productive response in a work environment like ours. We need them to be able to say, "Hey, what you're saying is unreasonable."
We've found there's a binary reaction to this approach: There are people who love it and find it's a really great way for them to grow. But we also find people who say, "This is not for me." People who don't want this kind of formal assessment tend not to last very long in our environment.
Besides testing someone's ability to say "no" effectively, what else do you look for?
Colin Gounden: In the final interview we actually give people some feedback. For a candidate we really like, we tell them something about their background or skill set that concerns us. We want to gauge their reaction: Do they get they defensive? Are they in denial? Do they get angry? Or, are they open and accepting of our feedback? We want to cultivate a culture where people feel comfortable saying, "I could have done that differently." People learn from their mistakes and they need to be able to acknowledge their mistakes.
What about motivating and developing the people you have now?
Colin Gounden: Most companies have a year-end review cycle which is completely ridiculous. We have created systems where we very quickly can give regular feedback to people; immediacy matters in feedback.
In addition to weekly informal check-ins, we have more formal skills assessments every 6 to 8 weeks. AI and blockchain are technologies that are moving incredibly quickly and we feel that our skills assessment and feedback process needs to keep pace to be effective.
(Note: VIA is not a client of mine, nor do I have any other business dealings with the company.)