Conference Day One: Tuesday, 28 August 2018
8:30 am - 9:00 am REGISTRATION AND COFFEE
9:00 am - 9:10 am Opening Remarks from the Chair
9:10 am - 9:50 am A Roadmap to Driving a Whole-of-Government AI led Transformation to Increase Efficiency, Reduce Operational Costs, Improve Service Delivery and Increase Citizen EngagementIan Oppermann - CEO, NSW Data and Analytics Centre
It is estimated that by 2020, 85% of customer interactions will be handled without a human agent. With the private sector amping up their AI led initiatives it is time for the public sector to follow suit. This session will bring together the leaders of public sector transformation, to devise a roadmap for a whole-of-government AI centred transformation. It will highlight both front end, customer/user facing applications and back-office applications. AI has the power to unlock even greater efficiency by taking automation to the next level.
- Identifying the main applications of AI in the public sector, from customer facing to backoffice
- Predicting and mapping up the pain-points that will delay the progress of AI technologies: Legacy systems, internal risk aversion, data cleansing amongst others
- Discussing the best model to drive this: Centre of Excellence, AI specific business group, A whole-of-government focused group
NSW Data and Analytics Centre
9:50 am - 10:30 am CASE STUDY: Creating an AI powered Portal to Help Plan the Future Growth of Auckland CityDr. Haydn Read - Head of Infrastructure Programs (DPO), Auckland Council
Auckland is in a period of massive planned growth with 10,000 hectares of land up for development. Most of these infrastructure projects will span generations and have a long term impact on the city. The Auckland City Council have created an AI powered portal that visualizes big data in a human consumable format. In one area of the portals functionality, it’s algorithms cut the green space into sections and are able to calculate the cost of development over time and optimize long term forward liability and explore some pressing questions for all cities, for example, intergenerational equity.
- Unlocking the Meta Data: Building partnerships with other council controlled authorities to create the lexicon for interoperability across assets and organisations
- Enabling next generation interrogation and insights tools that can be shared across organisations that create more accurate predictions, that are integrated across sectors
- Setting up big data visualization capabilities that allow it to be presented in a human consumable format
Dr. Haydn ReadHead of Infrastructure Programs (DPO)
10:30 am - 11:00 am Speed Networking
11:00 am - 11:30 am MORNING TEA
11:30 am - 12:10 pm Exploring the Opportunities of AI and Machine Learning in Healthcare for Improved Patient OutcomesDr. Brent Richards - Medical Director of Innovation, Gold Coast Health
IntelliHQ, a not for profit organisation promoting AI in Healthcare, has a unique set-up where it partners with Gold Coast Health, researchers, universities, private sector and start-ups to explore the opportunities of AI and boost implementation capability to make this a reality. On the agenda is creating streaming physiology data lakes, semi-automated recognition of ear drum pathology and investigating how to streamline patient referrals that could help save millions of dollars.
- Exploring AI and natural language processing to streamline patient referrals
- Partnering with researchers, start-ups and businesses to build capability
- Creating intensive care data lakes to identify new opportunities for AI in healthcare
Dr. Brent RichardsMedical Director of Innovation
Gold Coast Health
12:10 pm - 12:50 pm AI to Augment Fraud Management: How Inland Revenue have Increased Returns Assessment Coverage from 5% to 100%Paulo Gottgtroy - Chief Data Scientist, Inland Revenue New Zealand
Inland Revenue is the public service department of New Zealand charged with advising the government on tax policy, collecting and disbursing payments for social support programmes, and collecting tax. The have deployed augmented AI to the high volume task of returns assessment which is critical to fraud prevention and management. The algorithms are the key to the effectiveness of this program and it is highly dependent on data. This session will give insights into data collection, modeling techniques and creating algorithms to optimize fraud management.
- Data collection to understand customer behaviour and their implications for fraud
- Data modelling to create algorithms
- Understanding the case study: Applying AI to augment fraud management and replicating this for other high volume tasks
Paulo GottgtroyChief Data Scientist
Inland Revenue New Zealand
12:50 pm - 1:50 pm LUNCH
1:50 pm - 2:30 pm CASE STUDY: Continuous Improvement of Bots at Transport for NSW to Increase Customer Experience and Personalise EngagementDennis Chan - Manager, Customer Experience Technologies, Transport for NSW
This session discusses Transport for NSW’s deployment of their Chatbot RITA through multiple channels, including their website, Facebook Messenger, Twitter, Amazon Alexa and Google Home. Dennis Chan will discuss the hurdles they have overcome in this deployment of a Natural Language Processing Bot through multiple channels and will discuss their omnichannel bot strategy. Not only this, but Dennis will delve into the future of Chatbots and their continuous improvement model to provide Proactive Personalisation.
- Analyse the Chatbot life cycle and ensure you have a framework for continuous improvement
- Addressing the challenges that continuous improvement presents and how to overcome them
- Best practice and lessons learned that can be used in your Chatbot deployment
Dennis ChanManager, Customer Experience Technologies
Transport for NSW
2:30 pm - 3:10 pm Using Artificial Intelligence to Streamline Accounts Payable and Mail Management Saving 600 Hours a Week: Investigating Data Governance Strategy and Stakeholder EngagementRuth Edge - Team Leader, Corporate Information, Cardinia Shire Council
The Cardinia Shire Council is in a period of massive growth and the records management team are inundated with work. Paper based correspondence alone has seen a 50% increase in the past 4 years but the workforce has remain largely unchanged. The main reason for this is that they have been using AI software to assist with Accounts Payable and managing paper based mail. This session will highlight the journey, with a focus on getting stakeholder buy-in, change management and data governance strategy.
- Rethinking the data governance strategy
- Allaying the fears of job loss by focusing on value add
- Effective change management and stakeholder engagement
Ruth EdgeTeam Leader, Corporate Information
Cardinia Shire Council
3:10 pm - 3:40 pm AFTERNOON TEA
3:40 pm - 4:20 pm Unlocking the Power of AI by Consolidating Data from Different Systems, Technologies, Locations and AgenciesRohan Baxter - Senior Director, Data Science & Special Purpose Acquisition, Smarter Data, Australian Taxation Office
Data convergence is the key to the success of AI by enabling better insights and powering improved decision making. The challenge for the public sector is bringing together the data from different platforms, technologies, locations and agencies. Where do you get started, how do you ensure the data is up-to-date and ‘clean’. Find out more in this session.
- Creating a strategy to centralize all your data: analyzing data lake approach of colocating the maximum data footprint and computing power to minimize data movement
- Ensuring that the technology infrastructure can support it and if not, making a back-up plan
- Addressing cybersecurity and privacy issues through encryption
Rohan BaxterSenior Director, Data Science & Special Purpose Acquisition, Smarter Data
Australian Taxation Office
3:50 pm - 4:20 pm The Use of Machine Learning to Deliver Predictive Parking to Ease Congestion and Improve TrafficRoger Rooney - Senior Project Manager, smart Parking and Machine Learning, ACT Government
The Smart Parking program was launched 2 years ago in a bid to reduce congestion, travel times and drive more retail business whilst creating a better experience for citizens on the road. So far it has achieved 90% accuracy by monitoring 5 data categories and using a tree based ensemble learning model. This session will investigate the learning model, identifying scope for improvement and improving data.
- Identifying the key data categories that will determine occupancy and availability at any given time
- The impact on the city and local businesses
- Improving accuracy by accessing more data sets and better quality data
- Understanding the key insights into human behavior revealed through machine learning
Roger RooneySenior Project Manager, smart Parking and Machine Learning