What is a Data Scientist and how to find the right one
Big Data and Business
With all the talk about Big Data transforming business decisions, I investigated the possibility of hiring a data scientist. The first step was looking into what they do and now they do it.
Some of the responsibilities on job boards include:
- Building machine learning products
- Defining and Communicating strategy and planning to business owners
- Leading teams or a team of machine learning developers
- Developing production code and maintaining data pipelines
- Developing machine learning automation models
This creates a very valuable individual for an organization, but can you find these skills and experience with just one person. I know of very few individuals that could potentially meet these job requirements. This is more the model of the entire team with each member bringing their piece to the puzzle for a complete solution. This is much more realistic but also much more costly.
Software Developers verses Architects
Let’s take software developers and architects as an example. The developer provides specific skills matched with a set amount of years’ experience. An architect is often looked at with a much broader scope to include, jack of all trades, strategist, business requirement communicator, manager and general technology evangelist. And as you can imagine, there are a lot more software developer jobs than software architects. The managers and recruiters that are tasked with filling these roles have experience understanding the differences and how to find the right fit. It is no small challenge. The compounded challenge with hiring a good architect is knowing they can handle your specific challenges right out of the gate having not worked with them in the past.
The Data Scientist
The Data Scientist arena is no different as you’ll find the same challenges at the same levels. This means you will have difficulty hiring the right individual and high turnover trying to roll them out. A bigger problem facing the hiring teams are finding data scientists that can both understand coding and at some level complete portions of it themselves. All too often, mixing someone that sees writing code only as a necessary evil into a team of developers that can’t imagine doing anything else, creates one personality clash after another. This is a guaranteed drain on efficiency for any organization.
The only realistic resolution that we’ve found to these types of hiring hurdles are to build a well-balanced team. This starts with thoroughly defining requirements for each role to include rules of engagement for all team members. Those rules of engagement should define equally balanced check points and responsibilities. When a question comes up about who would do what, the team mechanics should be self-evident as to where each responsibility falls.
At Sparkfish, we focus on providing our clients with the full resources of an entire development team including data science, project management, UI/UX design but without the hiring hassles. We can provide the surgical precision needed on development and data projects to exceed expected results. Our teams run like well-oiled machines as they have been time-tested and worked out all the kinks that only come with experience.