HMRC boosts data scientist numbers


HM Revenue and Customs (HMRC) has significantly increased its spending on data science expertise, according to figures revealed by a Freedom of Information (FoI) request.

The FoI findings have been analysed by the Parliament Street think tank, who have been investigating the data capabilities of the department over the past three financial years.

The FoI request found that in the year to March 2022, HMRC employed 919 people with data in their job titles – such as data scientists or data analysts – with a total salary spend of £35.2m.

In 2023, that roster had climbed to 1,051 staff, with a total salary bill of £43.2m, while in the period to March 2024, data staff numbers had risen again to 1,279 employed in such roles, who between them created a total salary bill of £54.4m.

That means over the timeframe measured, HMRC increased the headcount of its data scientist and data analyst team by 39%, while HMRC’s data payroll jumped by 55%.

An HMRC spokesperson said: “Our data scientists play a key role helping us find innovative ways to better use the data we hold through technologies like artificial intelligence (AI) and machine learning.

“They provide a richer understanding of our customers and their needs, ultimately improving the services we offer to them. They also help us prioritise compliance work.”

Store of data

It’s perhaps no surprise that HMRC is building out its data science capabilities, because it has one of the largest stores of data on people and companies in the UK.

HMRC now has one of the largest groups of data scientists and AI experts in government who regularly contribute to cross-public sector initiatives, plus a wider community of staff across other areas of the department with an interest in data and upskilling themselves.

Data scientists use software, AI and machine learning to analyse and interpret large amounts of data. The government’s national career service suggests that salaries for this role can range from £32,000 for a beginner up to £82,500 for an experienced data scientist.

Scott Lewis, senior vice-president for Ataccama, said the government’s drive to recruit data scientists, against the backdrop of its investments in AI technology development and governance, shows it understands the need to apply best practices in its own operations.

“Advancing technology adoption, however, is a marathon, not a sprint, and requires stewardship, investment and oversight, and hiring people with the right skill set puts you merely at the starting line,” he said. “The challenge for the government will be to support their career development and role fulfilment in order to bolster retention.

“To ensure success, it will be critical to enable these people to do their jobs to their best ability – that means defining their role clearly, helping remove roadblocks, and providing ample opportunity for ongoing learning and upskilling as technology rapidly advances. We know that many data scientists spend a lot – sometimes most – of their time cleaning the data, a routine yet necessary task, before they can even start to work with it, create models and maintain it.”

Importance of AI

Lewis went on to highlight the importance of AI to maximise the value of the department’s new data staff, and said: “Utilising AI-based tools to automate manual, tedious tasks can accelerate data cleansing and transformation so that data scientists can get on with the more interesting, high-value work.

“This will also provide high-quality data to feed AI-based analysis and intelligence tools which will provide the government with accurate and useful insights to support better decision-making across the board,” he said. “AI may be the latest hyped technology, yet it is important to remember it is not the end goal itself, but rather the accelerator that will get us there faster.”

Computer Weekly recently reported how HM Treasury had been expanding its headcount of data scientists as part of a drive for data transformation across the department, which has seen it set up a new data management branch alongside formal roles in “advanced analytics”.

Richard Bovey, managing director at AND Digital, said: “Managing huge volumes of highly complex personal and financial data requires your staff to be equipped with the latest digital skills.

“Data scientists play a crucial role in helping large organisations unlock data silos and build tailored products,” he said. “They also own an increasingly important role in establishing and monitoring ethical usage of data. By doing so, data management builds trust in the organisation and facilitates stronger customer relationships.”

Skills shortage

Data scientists are in extremely short supply, and it’s a long-term problem. UK universities have only been able to supply around 10,000 data specialist graduates per year, but figures suggest the actual demand for staff with those skills could be 10 or 20 times higher.

However, on the skills side, there could be some good news: data released earlier this year showed that from April 2020 to March 2023, 7,600 students had enrolled on AI and data science postgraduate conversion courses.

Funding to the tune of £13.5m for these courses had been launched by the government in June 2020. While £3.5m of that was to assist with course costs, the bulk – £10m – was to deliver 1,000 scholarships aimed at women, Black students and disabled students, among other groups considered to be underrepresented in higher education. 

“Evidence suggests that the programme has had a substantial positive effect on the number of postgraduate students and graduates in AI and data science in the UK with expertise to address skills gaps in these industries,” said the analysis published by the Office for Students back in February this year.  

The programme is expected to deliver at least 6,000 new graduates in total, well beyond the target to deliver 2,500 by spring 2023. Many of these students had a first degree in a subject unrelated to AI or data science.

Between April 2020 and March 2023, 37 postgraduate conversion courses in AI and data science were funded and delivered across 28 universities: 30 courses were entirely new.



Source link