Skip to main content

The Data Science Accelerator is back

Posted by: , Posted on: - Categories: Data science, Data Science Accelerator

GDS have now run 5 rounds of the Data Science Accelerator, and several participants have since gone on to roles where they can use their new skills.

We are proud of this success. The Accelerator is a truly cross-government programme, backed by GDS, the Office for National Statistics, GO-Science, and the analytical professions (statistics, economics, operational research and social research). People who take part consistently tell us what great opportunities it gives them.

Our current cohort has expanded to Newport and Sheffield, bringing on board several new mentors and many new organisations such as Public Health England, DEFRA, Greater London Authority and the UK Hydrographic Office.

Now it’s your chance to apply - either as a mentor or a participant.

But why would you want to? What is in it for participants? You will learn more about the work of government and make great visualisations using code and statistics. You get an unlocked Mac (download any software you want… within reason!) while you’re on the programme, plus training and time to complete a useful, real-life data science project. Participants find that coding skills learned on the Accelerator help them do their day jobs better. Many of our products are proof-of-concept, but some have become regularly used tools.

Past analytical techniques include Machine learning, Natural language processing, social media analysis, geospatial tools, web scraping and various types of visualisation (e.g. R Shiny and Leaflet and D3 javascript libraries). But we’re always open to new ideas...

Weichao Wang describes the lessons he learnt from Twitter API rate limiting
Weichao Wang describes the lessons he learnt from Twitter API rate limiting

What happens on the Programme?
Each participant gets a laptop and a mentor. We ask you to commit at least 1 day per week for 3 months. We encourage you to spend that day either at GDS (Aviation House, London) or at the Newport and Sheffield hubs. At the end, you present your work to your home department and conferences organised by your profession.

How do we choose participants?
The most important things are a really good project idea (high value to your department and the public) and enthusiasm to learn. Coding experience is an advantage (e.g. Python, R, HTML / Javascript), but not essential. We’re aiming for about 10 people this time round.

How can I become a mentor?
We have welcomed 5 new mentors to the current round of the Accelerator and we are always eager to find more. We are looking for people with significant experience of doing data science projects and giving advice to others about data science.

You will meet up weekly with a participant, talk through their project and give them strategic advice (e.g. which of their potential questions has highest value) and practical input (e.g. helping debug code or suggesting a good library for them to use). You need to spend some time each week in the hub, though not necessarily a whole day. The coaching and help takes a few hours per week - any other time you’re located at the hub can be spent doing your own work.

Mentoring is great for your personal development (practising your coaching skills and getting experience of a wider range of really interesting projects). We don’t have a formal application process for mentors (though this guide explains a bit more), so if you’re interested please simply email!

How do I apply?
Please email this form to by Tuesday 31 May. I coach potential applicants and help with draft applications, so if you're thinking of applying but aren't sure, there's a real advantage in getting in touch. Here are FAQ and more information about what you’re signing yourself up for! We plan to start the programme on 30 June.

Sharing and comments

Share this page


  1. Comment by Derek Armstrong (Treasury Board, Government of Canada) posted on

    Pretty neat model for building a data science community. We're doing something similar here in Ottawa, though less formalized. We hire people in tranches at various levels (usually more junior), who have the coding skills and train them up on various pieces of government business. Then we send them off for promotions to departments that are clamouring for their services. We've discovered that the types of people engaged in data science tend to be curious about 'stuff' and that they naturally apply their data science tools to any new setting they land in. So far, so good!

    I literally laughed out loud when I read your hook: You get an unlocked Mac (download any software you want… within reason!). We face the same problem here, only IT keeps trying to wrestle our computers away from us.

  2. Comment by Ian Litton posted on

    Will you consider applications from local government?

    • Replies to Ian Litton>

      Comment by David Wilks posted on

      Yes, we are open to Local Government applications - please get in touch to talk about your project idea!