Service NSW wants to offer AI-powered services direct to citizens

Service NSW is focusing its early artificial intelligence efforts on improving the knowledge available to staff when interacting with the state’s citizens.

The agency, which acts as a single, whole-of-government point of contact, is taking a phased approach to its adoption of artificial intelligence technology.

Initially, it will focus on back-office applications, mostly around knowledge management, to enable staff to better serve customers, director of technology operations and delivery Colin Jones said.

In the longer term it hopes to provide AI-powered services directly to citizens.

This case study is part of a feature on the adoption of artificial intelligence within Australian enterprise. Download the report to read more individual case studies as well as analysis of the broader state of play.

“We’re still very early in the journey,” Jones said. “We’ve not implemented any of this yet, but we’re working on what the strategies will be for that.

“The point I need to make is that our core business and therefore our primary focus has to be our internal business, [which takes the form of] knowledge management for staff to service our customers.

“Once we’ve got those learnings and we’ve developed the knowledge management platform and whatever [AI] component [we choose] around it, we can then take those learnings and project that to customer-facing applications.”

Since its 2013 debut, Service NSW has expanded to cover 970 government transactions. It wants to transact 70 percent of all state government services digitally by 2019.

There is enormous diversity in those transactions and the agency wants to make sure its staff are equipped to deal with any and all requests that come across their desks efficiently, as per their remit.

“Because our core business deliverable is to be able to deliver whole-of-government service transactions for over 50 different agencies and 1000-plus different transaction types, our model needs to be as generalist as possible, otherwise it becomes cost-ineffective and you go back to the old world of speaking with lots of different government agencies [to transact with government],” Jones said.

“We need to be able to enable our staff as best as possible to be able to answer any questions and to be able to transact any type of business.

“But it’s not realistic to expect human beings to remember and be experts in all of those transactions, particularly when we expand the number of agencies and transactions we cover.

“We need to have really clever knowledge management systems to be able to surface very rapidly the exact knowledge article that we need, giving staff precise instruction that assists them on the transactional journey.”

AI may not be part of the knowledge management project from day one. While it is being factored into planning, it is likely to represent a later stage of maturity, and therefore a future enhancement to the knowledge management system.

A large number of AI and machine learning technology options are already being canvassed.

They include AI add-ons to Service NSW’s existing Genesys contact centre platform and Salesforce customer relationship management system.

Service NSW has also looked at IBM’s Watson “a few times” and – after recently outing itself as a trial user of the Google Cloud Platform – is also “thinking” about how it might use Google’s machine learning services in its approach.

In the future, Service NSW hopes AI will allow it to be more predictive and proactive in serving citizens – predicting what they might want or need, and then helping them make a start on the transaction.

“There are plenty of things we can do around AI and machine learning where we can suggest or predict what the customer wants based on their life journey with us, using their MyServiceNSW account, what questions they’ve been asking and what questions they’re asking this time,” Jones said.

“You know the customer, you know where they are in their life journey, you know what they’ve done previously, you know the likely transactions and interactions they might want to do.

“We can then help by surfacing the right information and starting the transaction for them.”

Download the report now