AI at work

From problem solving to business success

Joel Robbie started Nod as a way to solve a family problem. Now the fintech start-up has been listed on KPMG’s coveted FinTech 100 as a company to watch.

The multi-award winning business is at the forefront of using Artificial Intelligence (AI) to automate document generation and compliance for financial planners. The technology uses the data within documents to take over repetitive document creation and checking tasks previously done by humans.

“When my Dad was diagnosed with cancer, we refinanced the house to buy a restaurant and I watched as my family struggled for financial advice. My mission in life became to make sure the sorts of things my family struggled with never happened to another family,” Robbie says.

AI wasn’t something Robbie had even studied at UOW. He graduated with a Bachelor of Psychology in 2011.

“When I was working in psychology I was in the biopsychology stream – looking at statistics, with lots of data - and you become adept at triangulating data and making sense of it,” Robbie says.

“Nod actually started life as a marketplace business. If a consumer had a question we would find a financial planner who could answer it. I worked out that we had to automate financial documents to solve for access to advice. There was a need to create efficiency in the process and change the economics.

“In any business, you make a lot of decisions in your head about what should be in a document. At Nod we are using an AI platform that takes the decisions out of your head and automates them. The platform learns from historical advice documents, and from use of the platform, to generate them more quickly and efficiently.”

While Nod is servicing the financial advice industry presently, Robbie says the potential to use AI in a variety of industries is boundless.

“The underlying natural language processing and use of document data is very scalable and can be adapted to suit other industries and different types of documents,” he says.

“With Nod, we have built an intelligent and scalable piece of technology that can be deployed in a variety of new and compelling ways across different sectors – not just financial services.”

With that in mind, Robbie says Nod is already expanding on its AI platform automating documents in financial services, with the company also looking to deploy their technology to customers in legal and property services.

“I think the best-use case for any artificial intelligence platform or product or piece of technology, is always to take away things from work that are not where you add the most value,” he says.

“As a financial planner, where you add the most value is in talking to clients, making clients feel good about their financial situation, making sure they feel like their world is secure, and anything which takes away from that is a good place to start to think about where they can start to put in some sort of technology,” he says.

It’s not just in the business world that AI is being introduced on a bigger scale. Smartphone users have been able to access AI for a number of years through the “intelligent assistant” Siri, first launched by Apple in 2010.

Last year Amazon launched its virtual assistant, Alexa, onto the Australian market, four years after its initial release in the US.

Ryan Hunt is a senior product manager at Amazon and has been instrumental in ensuring Alexa’s success in the Australian market.

The UOW Computing Science graduate says his job involves not just interpreting and analysing data to support the AI platform but “teaching” Alexa about the Australian lexicon and culture.

“My previous job was at NewsCorp working on the innovation team where we were investigating how technology could help or hinder journalism,” he says.

“I was writing chatbots (a computer program which conducts a conversation via auditory or textual methods) for delivering news over voice – converting text to speech – and fell in love with the technology.

“One of the things I found is that people behave differently when they are talking to a robot especially the way in which they ask questions. People from different parts of the world, or even a country, may phrase a question differently. For example, one person may ask ‘How old is someone?’ while another may ask ‘What is someone’s age?’.

“There are a lot of cultural differences in each country in which Alexa has been launched and there are a series of steps that have to be followed with any voice product. It’s more than just being able to convert noises to words.

“In Australia we say things differently than in the US and the AI platform has to figure out the kinds of things it is being asked before it can figure out the answer. Then we have to ensure the platform also has neural linguistic understanding so the questions are being sent to the right area in Alexa’s ‘brain’. For example, if someone asks Alexa about JB HiFi I need to make sure Alexa knows it is a shop and not part of a stereo system.”

While AI is becoming more accessible to the general population, Hunt believes it needs more investment and smart minds working on it before it can reach its full potential.

“Any AI needs years of humans doing something over and again for it to recognise patterns and it will get to a point where it won’t need human intervention. But society is always changing – morals change, laws change, and it is hard for a machine to react quickly to these changes,” he says.

However, Hunt says the experience that Alexa now delivers is something that third parties can already build upon.

“It is like apps on phones, they come on to a platform and can provide other services,” he says.

“For example, there may be a company that maintains medical rules or laws which could then provide a voice experience for those in the medical profession where people could ask it about certain skills or policies. The opportunities available today for third party experiences are broad and rich.”

There is some uptake of these voice services already in play, says Hunt, with companies such as GoGet car sharing service, as well as for more general activities such as ordering pizza, and checking flight status.

“It requires businesses buying into it, with the experience and the level of AI used, driven by the businesses themselves” he says.

Joel Robbie
Bachelor of Psychology, 2011

Ryan Hunt
Bachelor of Computer Science (Secure Distributed Systems), 2001
Master of Business Administration, 2014