leader /'līdәr/ n: Anyone who holds herself or himself accountable for finding potential in people. - Brenè Brown
Leading Data Scientists
For me, it is constant learning how to lead a team of data scientists and how to be the kind of leader that I want to be. Because I have always worked with data, I don't want to step on other people's toes. Being too involved, micromanaging, often stifles creativity and unique solutions. Instead, as a leader I am accountable for finding and nurturing potential in people.
It is important to give people the opportunity to excel and to demonstrate skills. Recently, I gave a data scientist an opportunity to mentor a new employee (a new data scientist to our team). He rose to the occasion and many on our team, including my VP, was surprised. Why? Mentoring was a skill he was never given the opportunity to demonstrate.
Formulating The Problem
One of the most important things to do as a leader of data scientists is to help formulate the problem. Tom Kelly, author of "The Art of Innovation" writes that too often people come with the solution and not the problem. This is especially true with data science. Many times we are forced to take a step back and ask, "What are we trying to figure out?" or "What is the problem we are trying to solve?"
Most of what we do in data science ends up as a dead end. Failure, finding these dead ends, is a great way to learn. Without failure, experimentation, it is often hard to learn.
I often say, "It is amazing what can be accomplished if no one cares who gets credit." That being said, it is important to let your data scientists be the heroes. They are the ones that need to be presenting findings to clients and to executive management.