QuantHub Discusses Validating Data Science Skills with BigDataBeard - Podcast
This week, QuantHub’s CEO Matt Cowell and Chief Data Scientist Nathan Black met with Cory Minton, founder of Big Data Beard, a Data Science and Big Data media company, to discuss the ins and outs of hiring and developing Data Science talent and how QuantHub’s platform assists in validating Data Science skills.
You can listen to the 40-minute podcast by clicking here. Be sure to listen through to the end to find out fun personal facts such as what Nathan Black’s favorite “walk on a stage song” is and where Matt Cowell will be going on his next vacation (hint: it’s not Hawaii, not even close!).
No time to listen to the podcast? Here's a brief summary of the discussion.
On QuantHub’s “running definition of a Data Scientist”
Nathan Black referred to the Conway Model for Data Scientists, citing that the three skill areas of the model – Domain knowledge, Math/Stats , and IT skills – contribute to the ambiguous definition of what a Data Scientist is and does. He explained that these three very different skill areas represent an infinite continuum of talents. He added that people tend to define a Data Scientist with an emphasis on one of those three areas. In short, his answer is “it depends”.
Why is it so hard for so many companies to find “good” Data Scientists?
The role of a Data Scientist is a very cross-functional one. You don’t typically find the three skill sets of the Conway Model in one single person. Matt believes that by trying to merge all these skill sets into one person and not segmenting these roles and skills more, as is the case in the software industry with Product Managers, “We’re sort of creating our own problem”.
How QuantHub helps organizations find out if they have a Data Scientist candidate that is appropriate for their project
Companies are getting a lot of candidates seeking out this high paying position. Then they are manually doing the tech screening for all of these candidates. Most of this testing is heavily programmer-focused, rather than statistics or modeling-focused. Matt explained that QuantHub offers a much more comprehensive and scientific approach to testing for all the relevant Data Science skills such as data wrangling, data visualization, machine learning, and more than just programming.
QuantHub is essentially testing, “Do you understand the concepts of being a Data Scientist or Data Engineer?”.
How did QuantHub come to recognize the problem of hiring Data Scientist and come up with a solution?
Nathan explained that while he was working at QuantHub’s parent company StrategyWise and trying to recruit Data Scientists, he found that it took a very long time to go through all the steps to thoroughly vet candidates. At the time, StrategyWise was a startup and so there was a lot of pressure to take a few short cuts by hiring people from name brand companies and schools.
They ended up making a number of bad hires using this approach. In the process Nathan and some people at StrategyWise decided to create a solution to more efficiently and confidently vet Data Scientists, and so, QuantHub’s approach was created. Because of the great results that StrategyWise experienced with their “QuantHub” hiring process, StrategyWise clients started asking for access to the process and so the QuantHub platform was created.
On how costly a bad Data Science hire is
Matt explained that this cost of a bad hire estimate varies across the country, but that on average, it would be about $200,000. He added that the cost of making a bad Data Science hire could easily get into the 6 figures in terms of the cost, which would be catastrophic for a small company.
One company executive who Matt met recently lamented that he spent an entire week just doing tech interviews for Data Science hires. Matt added that companies like that one are spending $15,000 just assessing candidate skills alone. They simply can’t afford to have scarce technical and data resources pulled off a project to do skills assessments. So it gets very expensive if you make a mistake in hiring.
Other ways QuantHub is helping organizations
“We want to be a strategic partner in the Data Science and Data Engineering mission. We want to help companies build great teams,” Matt explained. QuantHub is not just a transactional partner in that it helps companies with the one hire. Matt added, “We want to help companies level up their Data Science skills internally as well, by using our assessments to benchmark skills in the organization.”
Matt and Nathan pointed out that there is a lot of Data Science training out there and too much choice – it’s overwhelming. So, people don’t know where to start with training to become a Data Scientist. QuantHub wants to become more of a matchmaker in that sense, helping would-be Data Scientists and Data Engineers to skill up in the most appropriate way. In that sense, QuantHub wants to help individuals and companies identify skills gaps and match employees to the right training using its skills assessments and progress tracking.
Is Data Science a “Team Sport”?
What does the future look like in terms of QuantHub being a platform Partner for clients?
Five years down the road for an organization who has gone through the journey of building a successful Data Science team and advanced analytics capabilities, “we really want them to look back on that journey and say that we were one of the most instrumental partners in them making that happen” Matt explained. “We want to be embedded in our client organizations and help them deliver on strategic priorities.”
Matt added that as far as the Data Scientist shortage is concerned, “We will be able to help the entire industry alleviate that Quantcrunch by helping people to really get into this field”.
How Big Data Beard would classify QuantHub as a “Matchmaker”
Big Data Beard suggested that if a traditional job board posting for a Data Science job is like Tinder swiping, then QuantHub is “e-harmony” of job hunting. You are actually matching people based on real experience. (laughs)
Can job seekers use QuantHub today?
Not yet, but that will be a future capability, probably coming next quarter. Data Scientists, Data Engineers and Analytics professionals can sign up today to be notified when this capability launches. And QuantHub won’t charge for candidates to sign up and take skills tests. In the meantime, QuantHub is running monthly skills challenges where people can sign up to take a 25 minute skills assessment for the chance to win prizes.
How does a company engage QuantHub?
The complexity of testing a Data Scientists really takes place behind the scenes in the QuantHub AI engine. For a company, vetting their Data Science and Data Engineering candidates using QuantHub involves a very quick sign up and discussions around roles and skill sets needed.
What does QuantHub think are the nearest term evolutions of the role of Data Scientist?
Nathan confirmed that the requirements are changing constantly. “We see the evolution of statistical methodologies in that people are not using the same statistical methods now that they were just 5 years ago.” He added that computing power is also changing the landscape in terms of how much you need to design your own algorithms versus just taking the latest cutting edge research and applying it.
Nathan explained that QuantHub must stay up to speed on the cutting edge of the evolution of Data Science skills, but also remain practical in terms of offering what drives value for companies. “We have to create a solution that directly mirrors job requirements of today while keeping up with shifts and trends,” he said.
How important is it for a Data Scientist to understand the evolution of cloud computing?
Nathan explained that it is highly critical to “some organizations”. The feedback from QuantHub clients has shown that there is a wide variety not only in terms of the definition of a Data Scientist but also what they need in terms of Data Science. Many companies may only need a power Excel user rather than someone who has cutting edge abilities with Spark clusters.
Advice from QuantHub for those considering a career or move to Data Science
Nathan explained that in his experience developing teams and in QuantHub that the shortest route is to take a top-down approach. He believes that at the end of the day a “Data Scientist is a just problem solver. They are just using data to solve the problems”. So he suggests that aspiring Data Scientists start with a problem they enjoy, and then go out and learn the components that it takes to solve that problem.
Matt added that he doesn’t necessarily think that you have to love or know all parts of the Data Science field in order to get into the field. You might not get into programming, but you might get into statistics and building models and analyzing business problems. He would suggest that you go into the field or areas of Data Science that you like because you can still make your way into these roles.
What it’s like to be a startup in Birmingham, Alabama
Nathan offered that on the one hand you are shielded from direct competitors by being in a different geographic space. “So you don’t see a competitor doing something and assume that you can’t do it too.” On the other hand, he added that QuantHub is always assuming that it is way behind Silicon Valley, which creates a kind of urgency to keep up with the rest of the nation. But then they will attend industry conferences and events and realize that the company is ahead of many.
Matt, who has been in Birmingham for 13 years and who grew another company there, believes that the customer-centric culture of the SouthEast is a huge advantage. He notes that there are great examples of companies such as SouthWest and Chick Fil A who are customer-centric. He believes that the customer-focused nature of the region is reflected in QuantHub’s approach as it tries to be very hands-on with organizations.
Matt added that Birmingham has a great startup ecosystem. Because it is smaller you can have more personal relationships with people that are creating businesses and having the same struggles, which is very valuable.
Finally, “there is actual capital in Alabama,” Matt says, noting that QuantHub easily raised $1.25 million in less than 30 days. That is not the norm says Matt. In places that are more crowded it takes much longer he says. He notes also that the low cost of living and labor contributes to a great startup situation.
Big Data Beard sums up QuantHub’s value proposition
“QuantHub has recognized that artificial intelligence which is powered by Data Scientists requires human intelligence first. And getting those intelligent humans in the right roles at the right time with the most efficiency feels like what QuantHub’s real value is to organizations.” 'nuf said. Thanks Cory!
Rapid Fire: Get to Know Matt and Nathan more! Let’s Get Personal!
Question: What is the last great book that you read that you would recommend?
Matt: “Monetizing Innovation”
Nathan: “Make It Stick” by Peter C. Brown (which he downloaded by accident looking for a different book)
Question: If you were going to walk in on stage at a conference, what would your “walk on” song be?
Nathan: A Queen song, probably “Don’t’ Stop Me Now” (wow)
Matt: “Welcome to the Jungle” – ( I’m a fan of the 80s and long hair)
Question: What piece of technology is currently making your life worse?
Matt: My phone’s alarm clock. I don’t like waking up.
Nathan: I was thinking PowerPoint at first. But really its “Internet Marketing”.
Question: What’s your biggest personal money pit right now?
Nathan: For Christmas I set up a Karaoke room inside our house. We are pouring money into it.
Matt: I would say music instruments. I end up buying more guitars because I feel like it will make me play more.
Question: What TV show are you binging on these days?
Matt: I love this series, it’s awful, I’m not gonna lie. It’s called Cobra Kai. The acting is straight up atrocious, but I can’t help it, I love it!” (Here's the link to the show in case you'd like to develop an addiction to Cobra Kai yourself)
Nathan: I usually spend most of my time watching tutorials if I’m choosing. My wife has me watch Outlander.
Question: What’s the next interesting trip you might be going on?
Nathan: Matt just informed me that I am going to Austin for QuantHub. My wife and I are going back to visit Japan where we lived for 7 years.
Matt: With college-aged kids we are doing a college tour to Michigan, Virginia and Penn State. (not exactly Hawaii!)
A special thanks to Cory Minton of Big Data Beard for the opportunity to talk all things Data Science, validated!