Job Questions

I’m recruiting at the moment. I’ve had quite a lot of people ask questions about the job and I thought rather than answer some questions for some people I would just answer all the questions on my blog and then point people here.

Before I do I will make a couple of observations about the recruitment process. Everything is done from the application form which is stripped of identifying information. We don’t look at CVs and I don’t take into account any previous correspondence with individuals when we’re marking the application form. I totally understand why people might want to say a bit about themselves when they get in touch to ask questions but do remember that we won’t remember any of it, or your CV, or any questions that you ask. We’re not allowed to do that and quite honestly I correspond with a lot of people and so I genuinely just won’t remember individuals. With all that said here are some questions and answers.

Q1. What are some of the main projects the team is currently working on, and what specific responsibilities would I have as a Data Scientist?

We’re quite a new team and at the moment all of our work is focused on a single project, which is delivering a demand and capacity model for acute hospitals, using simulation methods to predict acute hospital activity over the next 20 years.

We use R, Python, Shiny, git, GitHub, and Azure in production and have access to a Posit Connect server to deploy our outputs.

Part of the reason for the expansion of the team is to give us capacity to move into other projects. One particular area of interest for us is text and we have some interest from other analytical teams inside the Strategy Unit to work on some projects to do with text.

Q2. Can the job be done hybrid/ remote?

The job is fully remote and we have team members all the way from London to the North East- equally there is an office in Birmingham which you can use to work if you prefer to do so. The only requirement is that you self fund travel to our away day in Birmingham four times a year.

Q3. How is the data science team structured, and what are the opportunities for collaboration and mentorship within the team?

We’re only a team of 6 at the moment but we have collaborative interests with other analytical teams at the Strategy Unit. There is a wide range of specialisms within data science in the team and you will be working alongside individuals with specialist skills who will be happy to share their learning and to share in your own skills and learning. The team culture is focused on learning, continuous development of best practice, and maximising opportunities for the team to work together/ minimising bus factor.

Q4. How often do data scientists interact with clients, and what is the nature of these interactions?

Most of our clients are internal to the unit, for example there is a team who manages the rollout of the model and most of our day to day interactions are with them rather than the endusers. Having said that there are opportunities to work with the endusers developing products and speccing out what they might find useful.

Q5. How does the team balance the need for innovative data science solutions with the practical requirements and constraints of the job?

We do whatever we need to do to get the job done. If we need an elaborate solution involving Spark and Docker, we’ll build it. If we just want a Jupyter notebook, we’ll do that. We avoid innovation for the sake of it at all costs, because it just reduces bus factor and leads to bugs.

Q6. Being the lead data scientist, what is your vision for the data science team’s growth and impact over the next few years?

The team are engaged in a highly complex and business critical modelling process for a national programme of work. At the same time they are situated in the Strategy Unit alongside other teams with lots of different types of analytical expertise and an interest in using and adapting data science methods in their own work. I’m building a team who are technically expert enough to do the sophisticated modelling work required, a team that is resilient and shares skills across team members to avoid single points of failure, and a team that understands the needs of non data scientists and is able to collaboratively build useful stuff with a wide range of clients, including analysts and non analysts. We open source all the code we can and we participate in communities like NHS-R and run training sessions to share our learning with the wider health and care sector. As a team we value kindness, honesty, and team spirit over technical ability.

Q7. I see there is formal training and hands-on experience could you elaborate a bit further on that?

The pace of learning in the team is pretty significant and each team member is given the training, time, and support they need to learn new skills. Sometimes that’s just spending a bit of time with one of the team, or it might be a training course or a conference. Learning is vital in the role from day one and we have a culture where learning is given its due prominence.

Q8. Given the emphasis on open-source work, how does the team decide which projects or code to contribute to the wider community?

We open source all of the code that we can and are currently negotiating with a client to release our modelling code as open source, which will be a significant contribution to others working in the area. Some team members have their own open source projects and the team are encouraged to dedicate enough time to these to keep them active alongside their other work.

Q9. Are the CVs and cover letters personally reviewed by a human?

Yes, yes, a thousand times yes! Never say never but I personally would not apply for any job where the materials are not reviewed by a human (although see the note about CVs in the preamble to this)

Q10. Interview Stages: How many stages are there in the interview process, and what do they entail?

It’s just one interview which can be conducted on Teams.

You will be asked to bring some code that you wrote and talk to us about it- how it works, what it does, how it could be improved, that kind of thing. Then just some interview questions about your skills and experience.

Q11. Will there be any technical assessments or case studies, and if so, what areas will they focus on?

No, I don’t believe in doing this. It may work for some, I’m not saying nobody should do it I have no idea but for me it’s not something I want to do.

Q12. Preparation: Is there anything specific I should prepare or review prior to the interview?

I’m looking for data scientists who can contribute usefully to repos in R and/ or Python that are in production, data scientists who are ready to learn new stuff and take the initiative with learning, and data scientists who want to work on a team, learn from others and contribute to the learning of others. I don’t worry too much about can you do this thing, can you do that thing, because we can teach that.