Recruiting for digital: look for passion and quality questioning, says Kubrick's Simon Walker
Founder Simon Walker on Kubrick's novel methods of recognising and supporting tech talent
Like the poor, it seems the IT skills gap will always be with us. Indeed, with many digital jobs, a lack of money is one of the hurdles. Many people simply can't afford a hefty loan and absence from work in order to take the risky step of retraining in digital. Others may not know where to start, or feel they are insufficiently technical.
Simon Walker, a film studies dropout turned entrepreneur, saw this mismatch first-hand when his father, a skilled crane technician but with no formal education, struggled to find a way to switch into the new economy. With a long experience in recruitment, including founding two startups with business partner Tim Smeaton, Walker also saw the need from the business side of things - it was very hard to recruit digital skills, and yet at the same time internal training was an afterthought.
In 2016, Walker and Smeaton came up with a way to square this circle - plus Walker had a promising name for the new venture from his university film days. They decided to take a non-traditional approach, focusing on one particular area - data engineering - looking at new ways to mine seams of talent, rather than headhunting existing practitioners. Thus, Kubrick was born.
The decision to focus on data engineering was not easy, said Walker, now managing partner at Kubrick.
"It was nerve-wracking, not just with the normal fears around starting a business, but everyone told us to do data science."
The focus on the nuts and bolts of data engineering rather than the rock-star realm of data science (which was at the time billed as the 'sexiest job in the world') came about because the big, messy organisations they had been working with were still struggling with the fundamentals. Data was siloed, dirty, duplicated, outdated and incomplete, and the organisations lacked a strategy to get it in order.
Opening doors to digital
A pivot to data engineering may not be too much of a stretch for a recent maths, physics or computer science graduate, maybe - although given the pace of change even older graduates in those disciplines may struggle - but many people who would naturally be suited for this job, such as Walker's father, find the doors permanently closed. It can be hard to find a way into a niche field.
Therefore, training became a key part of the offering. Kubrick's model sees it paying budding data engineers for 15 weeks as it trains them in the latest tools and skills - including 'soft skills' - before placing them with its customers during which it continues to pay them a salary. After two years, they are free to join the company as a full-time employee, no strings attached.
The model has proved successful, both in pulling in new blood and also satisfying the needs of its clients, Walker said. In an internal survey, 96 per cent of candidates who found placements through the scheme said they would not otherwise have got a job in data. However, there is scope to expand the pool: most candidates thus far have been recent graduates - albeit in disciplines ranging from biochemistry to history.
In particular, the well-documented problem with of bias in AI calls for more data engineers from diverse backgrounds, people who can ask different questions.
"You can't have a bunch of people that look the same sound the same, have the same age group have the same gender because they will not ask the full set of questions that you need from that data," said Walker.
So, a continuing focus is to build communities via podcasts such as Black Tech Talks, which features black leaders in technology and data into talking about their experiences, through social media drives and better targeting.
"We did a lot around building those groups, being able to invite those people in, and we spent a lot of time on understanding where do we advertise? Who do we talk to? What affiliations do we have? That way we can start attracting those people that wouldn't have looked for a career in data."
Where's the passion?
A key, but elusive, metric for success in any field is passion. A talented computer scientist may make a lousy data engineer because they have no interest in the plumbing; conversely, someone from an entirely unrelated background might find that all the pieces slot naturally into place for them.
Evaluating a candidate's passion for a topic is a human skill, but there are ways to assist in this judgement, Walker said, including adding small tests in the training process and seeing how candidates approach the training, although obviously this needs to be done with sensitivity and transparency.
"The advantage of being a data company is you have loads of people that are able to run these experiments. We think we could maybe open a candidate tool that might show up our blind spots, and actually might get them into data if it's done in a really non-intimidating way, and they can do it anonymously."
All change
The connected fields of data engineering and machine learning are changing rapidly and of the original 2017 data engineering syllabus only about 35 per cent remains, the rest having been updated to fit current needs. Candidates that go through Kubrick's training processes also retain access to the updated course materials, the idea being that this can help them stay current in what tend to be very short-tenure positions.
The original data engineering focus has been expanded to include cloud management and AI, and the company has grown significantly, to over 1,000 employees, but the supported training approach remains, as a way into the digital sphere for those that would, for whatever reason, be excluded, and to meet the needs of UK Plc.
The most sought after digital talents, according to a survey of its corporate customers, are communications skills and people the ability to ask the right questions, neither of which technical.
"We think often you can define the quality of a data engineer or a machine learning engineer, by looking at the quality of that questioning, so why are they doing something. It's not tech for tech sake," Waker said.
Digital transformation calls for a range of skills and personality's and Kubrick's model can help to provide them, he added: "We work withreally industrious people, future generations, junior professionals, and toretool them and give them a skill set that will not only future-proof their careers, but it also would help industries and society as a whole."
You may also like
/feature/4334521/tech-isnt-meritocratic
Leadership
Tech isn't as meritocratic as you think
And relying on graduates to fill vacancies isn’t working
/feature/4331535/long-reads-women-experience-imposter-syndrome
Leadership
Long reads: Why do so many women experience imposter syndrome?
And is it always a bad thing?
/sponsored/4323596/case-study-triad-group-transforms-dfts-renewable-fuel-obligation
Cloud and Infrastructure
Case study: Triad Group transforms DfT's renewable fuel obligation
A simple, reusable set of components