Choosing a company to intern with is always a daunting experience. As anyone who’s ever completed a commercial internship as part of their PhD will know, it’s difficult to decide which company is likely to offer the right mixture of data science training, commercial experience, and induction into everyday working life.
Gregory Barbour, a Research Intern in Faculty’s R&D team, tells us about his day-to-day work on our AI privacy project, why he was drawn to apply for the Faculty internship, and how it’s setting him up for a career in data science.
First thing’s first; how did you land your internship at Faculty?
I’m in the third year of my Data Intensive Science (DIS) course at UCL; during that year, we’re encouraged to apply for six-month commercial internships. By the time I had to start applying, I already had a pretty clear idea of what I wanted from my host company. I wanted to put my data science skills into practice and see them making an impact in real-life situations, but I didn’t want to lose the things I love about academia. I wanted to still have those opportunities to conduct research, test new ideas, and keep fine-tuning my data science skills.
I had been enthusiastic about Faculty’s work from Tom Begley (Faculty’s R&D Lead) lecturing on the world of AI consultancy, the variety of projects Faculty have tackled, and the impact of today’s commercial AI. Faculty’s teaching really stood out; you could tell that they had a lot of experience in making data science work accessible to PhD students, and in turning PhD students into data scientists through their fellowship programme. I was also really excited by the variety and impact of their work.
I spoke to Tom about the possibility of working with him and his team, applied for the internship, and then landed a place. For the past couple of months, I’ve been deeply embedded in Faculty’s R&D team, working on guaranteeing privacy in data science.
How have you found the transition from academia to life at Faculty?
My first weeks at Faculty were spent reading papers, learning the framework and scope of work in data privacy. It immediately felt very similar to academia; much of the research in this field is done by academics as well as private companies, so the methodologies we use to build, test, and roll out solutions are very similar.
I met the R&D team, who are all former or current PhD students like myself. Tom, alongside Chris Frye, are the senior members of the team. They’re experts in AI, data science, and AI safety, so they’re hugely involved in all the projects in R&D.
The R&D team is mostly made up of Faculty data scientists who have been seconded to it from the other, more commercially-focused side of the business. The idea behind this is that the commercial business should stay in close contact with cutting-edge research that drives innovation.
The R&D team isn’t here to just publish papers and look clever. Everyone is hyper-focused the impact of our innovations for customers. For example, a number of clients are already very interested in seeing (and using) the results of our research into privacy, and our explainability work has found its way into some major commercial projects.
Have you found it valuable working alongside data science specialists?
Meeting and working with so many dedicated data scientists – all with firm backgrounds in physics, maths and science – has been a hugely enriching experience. I’ve learnt new techniques and tools for data science, and I’ve also gained some valuable insights about working as a data scientist in industry.
Having so many data science experts around also means that I can call on a huge range of expert knowledge at any time. Faculty is a very collaborative environment; whether it’s a simple tech issue, or a maths problem in a project that needs three physicists and a whiteboard, everyone is willing – and able – to help out.
Meeting and working with these people has been perhaps the most enjoyable part of my time here. It’s given me invaluable inside experience that I can take forward into a future career.
What kind of projects have you come across in your time at Faculty?
I applied to intern at Faculty partly because of the variety of projects available – they’re really dedicated to making AI work for everyone, in a huge variety of fields. As a result, data scientists at Faculty work across a broad variety of sectors: I’ve talked to people who have helped the government root out potential terrorist propaganda, built trading models for financial services companies and helped rugby teams find the best talent.
It is a very different environment to working for an in-house data science team, where I’d expect our work to be closely focused on a relatively small range of use cases.
It’s been really exciting – and educational – to see first-hand how the application of data science can benefit so many fields.
How would you summarise your time at Faculty?
My experiences at Faculty have been very positive. I’ve already learned a huge amount about new data science techniques, about AI safety, and about how AI is applied and used in business. I hope to continue along this path in the coming months, and take my experience in this internship forward into any future career.