Working with data: It’s more human-centred than you think
From left: Mina Dashti, Anne Sloman, Oonagh O’Shea and Diarmuid Cahalane take part in a panel discussion at Accenture’s AI, Data Science & Big Data Open House. Image: Luke Maxwell

Working with data: It’s more human-centred than you think

27 Oct 2017237 Shares

At Accenture’s Open House event, we learned that working with data is as much about dealing with people and putting them at the centre of your decisions as it is about numbers and artificial intelligence.

AI, data science and big data – all buzzing terms in the sci-tech space and sometimes victim to the hype around them.

At Accenture’s AI, Data Science & Big Data Open House event, discussion on these terms dispensed with futuristic dreamscapes in favour of the real and everyday applications taking place right now. A panel of practitioners who have been working with data analytics for years (before it was cool) shared their practical advice for dealing with clients and solving problems with data, revealing the more human side to the roles involved.

Accenture Analytics in focus

Paul Pierotti opened up proceedings with an introduction to the various analytics teams working at Accenture. Pierotti leads them all as head of Accenture Analytics, and was nothing short of animated in an evening showcasing the work they do.

In all, Accenture’s analytics teams span four categories. There’s the Tech Labs team of experts in artificial intelligence (AI), then the Advanced Analytics team who exhibited the healthcare analytics they are using to transform patient outcomes.

The Analytics Delivery team, Pierotti explained, is all about “delivering analytics insights at scale”.

“They blend the decision science and the computer scientist expertise so that we can actually get those wonderful insights and implement them across large clients,” said Pierotti.

The final team highlighted on the night were those at the host building, The Dock, Accenture’s all-new research and innovation hub in Dublin’s docklands. “They’re actually working, alongside with the Tech Labs, to do new artificial intelligence and other related solutions, and implement them in our next-generation services,” said Pierotti.

Start with the problem, not the data

Reflecting on the scope of individuals and skillsets that make up Accenture’s analytics business, Oonagh O’Shea from Accenture Digital said: “It’s really a team effort, it’s not a one-man job.”

O’Shea, who is a senior manager responsible for advanced analytics in life sciences, participated in the panel discussion, and her sage advice for any team about to tackle a problem using data was not to start with the data, but the problem itself.

“Obviously data is fundamental and data is really important, and we’re obviously going to use that and use the immense amount of data that’s being created on a daily basis, but it’s really [about] getting it back to the business. How is it going to solve a problem for the business?” she said.

This advice might come as a surprise to clients who want to dive into data without first knowing what they want to achieve. Mina Dashti, an analytics manager at The Dock, faces a similar reality check with clients who want to apply AI as some kind of problem-solving wand.

‘Sometimes clients think that AI is the magic thing that will solve everything for them. No, it’s not that. AI solves the right problem with the right set of data’
– MINA DASHTI

“They only know that they have a problem but they don’t know what is that problem and if that problem is solvable, and to what extent it’s solvable with AI. Sometimes they come and they think that AI is the magic thing that will solve everything for them. No, it’s not that. AI solves the right problem with the right set of data,” said Dashti.

Human-centred data analytics

Diarmuid Cahalane, who works at The Dock as a lead research scientist teasing out cutting-edge AI technologies, was happy to see the panel discussion putting people at the centre of data problems and problem-solving.

“I think we got into a couple of interesting human-centred problems. How do people work with AI and analytics, how do we achieve the best value from those technologies, and then the societal implications: how do we understand and explain what the technologies are doing, how do we manage transparency around our data,” he said.

Transparency was certainly part of the subject matter tackled by the final panellist, Anne Sloman, a fresh face at Accenture having joined the team this summer. Sloman has 14 years’ experience in retail analytics and, lately, her clients are largely concerned with tackling one big game-changer: the incoming EU General Data Protection Regulation (GDPR).

“It’s actually something I’m quite passionate about,” said Sloman, who was happy to take questions on what will happen when this new regulation comes into effect on 25 May 2018.

“I do think it’s a real challenge for retailers. They have large customer bases that they will have signed up a long time ago, they have lots of different legacy systems, so there is quite a challenge in terms of how they might interpret the rules.”

Overall, the evening at The Dock was a glimpse through a window on the world of working in analytics, which certainly involves mathematics, statistics, data modelling and visualisation, but also a hearty helping of human-centred approaches, strategy, communication and problem-solving.

“It’s nice to get that feedback from people that they’re interested in hearing about how we work with clients, not just about all the machine learning and data science and all the more technical stuff that we do,” said Sloman.

Elaine Burke
By Elaine Burke

Elaine Burke is managing editor of Siliconrepublic.com. She joined in 2011 as a journalist covering gadgets, new media and tech jobs news. She comes from a background in publishing and is known for being particularly persnickety when it comes to spelling and grammar – earning her the nickname, Critical Red Pen. When she hasn’t got her nose stuck in her laptop, you’ll find her in the kitchen, at the cinema, or on the dancefloor.

Loading now, one moment please! Loading