“Data scientist” is an on-the-rise job title, but what are the skills that make a good one? And how can both data scientists and the companies they work for make sure data-driven insights become actionable?
In a recent interview, DJ Patil (@dpatil), formerly the chief scientist at LinkedIn and now the data scientist in residence at Greylock Partners, discussed common data scientist traits and the challenges that those in the profession face getting their work onto company roadmaps.
Highlights from the interview (below) included:
- What makes a good data scientist? According to Patil, the number one trait of data scientists is “a passion for really getting to an answer.” This does mean, Patil said, that personality might trump skills. Pointing to what he calls “data jujitsu” — the art of turning data into products — he noted that some people can approach a problem “very heavily and very aggressively” using all sorts of computing tools. “But one data scientist who’s clever can get results far faster. And typically in a business situation, that’s going to have better payoff.” Patil pointed to a site like Kaggle, where people compete to solve data problems, and noted that despite the number of data scientists there using machine learning and artificial intelligence, some of them are “getting beat by people who just have good, interesting insights.” [Discussed at the 1:34 mark.]
- Despite the “data smarts and street smarts” that Patil sees as key to data science, data scientists sometimes struggle to get companies to pay attention to the insights data science can provide. The good news is that Patil anticipates this attention issue will fade in the future. Once organizations recognize the importance of data, they’ll identify and handle data in better ways. Furthermore, we’ll see “a new generation of designers, product managers and GMs who are also data scientists and not just former engineers.” [Discussed at 4:09.]
The full interview is available in the following video: