X
Business

Aussie developers better for web than big-data business: R&D head

A proclivity for teaching web-design skills over hard-core data science will leave Australia struggling to meet rapidly increasing demand for business intelligence and big-data specialists, the local head of R&D with Dell Software has warned.
Written by David Braue, Contributor

"We have to go through a lot of people before we find the right one," explained Guy Harrison, Dell Software's Melbourne-based executive director of R&D for information management.

"We go through Java programmer after Java programmer, and find all they really know how to do is to create a website. Web development is almost like join the dots — but for commercial software development, you want people who are able to solve problems in new domains."

Guy Harrison
Image: Guy Harrison

That's made building up Dell Software's local brains trust a constant challenge for the Melbourne-based R&D team that Harrison heads, which joined the growing Dell Software family with the company's US$2.4 billion purchase of business-software giant Quest Software late last year.

Quest's ongoing R&D investment — which centres around database, business intelligence, and big-data development — "is not very common these days with larger software companies", Dell Software ANZ managing director Ian Hodge said, but it has become more important "as we move from what was effectively a hardware company that everyone knows into a US$1.5 billion software business and an end-to-end solutions company".

To support this goal, Harrison wants to build an Australian team filled with problem solvers — but finds that most programmers coming from local universities are not maths-oriented, and "end up doing things they think are more sexy, like games development, or softer skills that are guaranteed to lead to a relatively easy job, like web development".

Universities could raise the bar by introducing undergraduate teaching or postgraduate qualifications such as the Masters of Data Science recently introduced at American universities such as the Illinois Institute of Technology — but he acknowledged that there is a chicken-and-egg problem, because few students will pursue the field unless they feel there is a guarantee of relevant jobs.

"You're taking a bit of a leap of faith at the moment, if you're trained as a data scientist, that the explosion of data science-oriented jobs is actually going to happen," he says. "The next stage of business success is going to be harvesting data with smarter algorithms to get ahead of other companies."

That's not a gamble that many people will be willing to take until the big data and business analytics fields improve their general branding and recognition. However, with analytics roles quickly gaining primacy within software development organisations, in many cases, the opportunities are already there.

Companies will increasingly value developers' analytics skills as they move to embrace maths as a competitive weapon, Harrison added, noting the particular challenges faced by retailers to meet new competitive threats.

"Globalisation and the internet make it hard to always be the lowest-priced vendor," he said. "You need to be the one that gets people's attention first, that offers them the unique combination of products that they want, that dynamically prices to match their circumstances."

Most industries are finding new value in big-data applications like those the Dell Software team is focused on helping design. Over time, Harrison wants to reinforce the R&D facility's position as a global centre of competency in the development of data-driven analytics that he sees as core to customers' business strategies — and Dell Software's success — in the long term.

"In every industry, there's some advantage of algorithms," he explained, "whether having better churn detection, better advertising, better medical outcome prediction. There isn't a great deal of software to help you get this done — but the smartest ones are doing it through brute force, and brain force."

"You want to get the smartest people in the world to build your algorithms, and one of my groups is working to create the software that makes it easier to build up those algorithms for advanced machine learning. So I believe those are the sorts of skills we should be building out of our universities — but I don't know if everybody has totally bought into it yet."

Editorial standards