Joanna McKenzie is our Principal Data Scientist in the Data Science team. She writes about the flexibility required for working as a data scientist and also managing the team.
Running the Data Lab’s data science team is a busy and varied role; it is definitely true that no two days are alike! The team delivers projects for clients from across the Scottish economy. These are data projects, and for the most part we limit our projects to 20 days of data scientist time. Our projects can be from a private sector, public sector or third sector organisation, and we work with all kinds of data in lots of different formats. We need to be flexible in our tools and approaches and to be ready to learn new skills.
I’m not a unicorn, but I do know how to solve technical problems
Since moving into management, I’ve found that most of my projects are now mentoring projects. In these, I become a technically-knowledgeable sounding board and facilitator for a person or organisation attempting something data-related for the first time, or otherwise stepping out of their comfort zone. Since I’m not a unicorn I don’t know everything there is to know about data science, and often these people are using tools, processes and datasets I’m not necessarily familiar with, and certainly not an expert in. That said, I do know how to solve technical problems, and often the key to these things is to make suggestions for what investigations might result in a solution. Tackling those problems themselves can really help the mentee gain skills and confidence in their more general data science skillset.
Pipeline management working at the Data Lab is a whole challenge on its own. Whereas a data science team in a large organisation might be predominantly working with the organisation’s own data and in their own sector, the Data Lab has a remit that spans the entire Scottish economy. It’s not uncommon for me to be talking about air travel in one meeting, healthcare in the next and computer game design in another. Often the aim of the conversation is for me to understand what gets them excited – what could I conceivably deliver with the data they have that would make a material difference to their organisation? I then have to pair up what they want to achieve with the data they have to identify a plausible but valuable project.
Our remit spans the entire Scottish economy
The Data Lab’s new service supporting organisations in their external funding bids is another space where I do a lot of pipeline management, and it brings in some additional challenges. Just as for my more usual projects, these can come from any sector of the Scottish economy. However since there’s grant money up for grabs these are often much more ambitious projects which require alignment with other professionals: researchers, project managers, and perhaps other experts. The scope of these projects can be both exciting and daunting, especially since there’s always the question of whether or not each project will achieve the grant money it’s seeking.
I was not trained as a data scientist originally; I gained most of my knowledge through my physics PhD and then later through many years working as analyst, senior analyst and team leader in the wind industry. I’m gaining confidence in navigating the differences between an analyst team and a data science team – for me the real difference is that an analyst does the same calculation many times in different contexts, whereas for a data scientist every calculation is new. That does mean that to do data science well you need to be an exceptionally good listener – your job is to invent a wholly new analysis process that your client can use to create value, and you need to be really sure you understand where the value is coming from.
A real joy of working at the Data Lab for me is being able to get involved in some of the exciting things the rest of the organisation are doing. I particularly enjoy those occasions when I get to speak to or work with our MSc cohort of data-scientists-in-training. I’m also looking forward to the first data science collaboration with Torch, which I feel sure will happen. It’s a very positive and vibrant place to work, and one thing is for sure: you never know what new challenge is just around the corner.
We’re looking to hire a Data Scientist to join our team: view the job description. Closing date 8 January 2021.
