How Energy Queensland is using generative AI to transform customer billing

The use of generative AI for CX and service is often focused on streamlining communications and operational efficiency – but it is capable of much more.

Ahead of his session at All Access: The AI Revolution APAC, Michael Dart, chief customer officer for Energy Queensland, explains how generative AI is being used to revolutionize the energy industry by orchestrating supplies with the aim of eliminating volatility and allowing customers to consume energy on a monthly fixed subscription price, as they do with a mobile phone or broadband contract.

CX Network: During All Access: The AI revolution APAC, you talk about Energy Queensland’s AI journey. Without giving too much away, tell us about some of your work and outcomes to date?

Michael Dart: We love all things energy at Energy Queensland (EQL). Our businesses are responsible for managing the supply of electricity to 2.3 million customer connections and for selling electricity to more than 760,000 retail customers, so we want to crack open the benefits of AI for our employees and our customers while managing risk, cyber security and protecting our reputation and brand.

As a taster, energy is bought and sold in a traded stock market environment the same way other commodities are traded in financial markets.

We use AI tools to predict the optimal time to charge our distributed batteries across the state and discharge them, to either profit from energy arbitrage or manage voltage volatility to prevent incidents within our communities where the grid cannot keep up with demand.

The complexity of the algorithms to optimize this in sub-second times and respond is managed autonomously via our AI tools.

We call this “orchestration” in the energy industry, and it is transformational.

One of the opportunities with solar and wind as energy sources is that they work best when the sun is shining, and the wind is blowing. Having the technology that orchestrates energy assets allows us to capture, store and re-use energy when our customers need it.
It also allows us to manage the demand side, shifting load like hot water, pool pumps and electric vehicle charging to optimize the available energy mix, keep the lights on and the cost of energy low for everyone.

RELATED CONTENT: Using sentiment analysis to drive customer satisfaction at Toyota Finance New Zealand

CX Network: How is Energy Queensland using AI in its contact center and what impact have you seen on customer metrics such as NPS and CSAT?

Michael Dart: The target is low value and low risk tasks. We want our talented people to be using their skills and experience on the most challenging and fulfilling activities.

It’s a work in progress as we develop a policy framework that is safe and secure and allows our people in the field who have been tasked with customer work and those on the phones and screens in our call centers to have confidence that the jobs AI are doing free them up to do the high-risk, high-value tasks that really improve a customer’s experience.

A great customer-facing use case for generative AI is using AI to draft responses to unstructured customer written communication like emails, SMS and letters, etc. Leveraging a natural language engine, data insights and generative AI, we hope to be able to draft responses to up to 60 percent of all non-voice inbound traffic. We don’t think AI will write better or more insightful communications than our skilled staff, but we do expect it will shorten our response times significantly. I’ve personally watched emails and agent response lines appearing on the screen while the customer is on the phone. What’s even better is the AI learns to write the email in the agent’s “voice”. Our research indicates that increasing the timeliness of our response nets around a 20 precent improvement in customer satisfaction on the engagement.

The secondary benefit is that we have created new capacity within our workforce which can now be re-deployed into higher value customer work which in turn has the potential to be a multiplier for CSAT and Net Trust Scores (NTS) or Net Promoter Scores (NPS).

RELATED CONTENT: Powering better CX and ROI through Voice of the Customer

CX Network: In your opinion, where in the customer journey can generative AI be most effective in boosting customer loyalty and retention?

Michael Dart: It’s taken us 110 years to build and operate the energy system in Australia today. The poles and wires that crisscross Queensland will still connect customers but what they’re connected to will be very different in the next quarter of a century.

We are going to increase the total capacity by 2.5 times to replace our carbon intensive baseload generation with a mix of sustainable and renewable energy sources, 90 percent of which do not exist today.

In an energy industry context, the complexity of the supply chain is practically unfathomable for our customers, and this is where generative AI offers the most potential.

We are going to witness creative destruction not seen since the industrial revolution.
The power of generative AI to orchestrate emerging sustainable energy technology like rooftop solar, home batteries and EVs with vehicle to grid (V2G) capability will enable EQL to re-engineer the way our customers experience energy. It’s entirely feasible that if we can manage the supply side volatility, in future our customers could be consuming energy with subscription service on a monthly fixed price, no different to your mobile phone or NBN. The magic will be creating a “set and forget” product for customers where they can sit back and relax and enjoy the savings.

RELATED CONTENT: What is a chief artificial intelligence officer – and do you need one?

CX Network: What are your three top tips for CX practitioners who are looking for an AI vendor to work with?

Michael Dart: In the coming AI era, data becomes one of your biggest assets.
You need to simplify your technology stack, get control of your data and where you can, avoid locking it down behind software as a service (SaaS) or cloud storage walls.

Prudent investment and preparation in your technology today will accelerate your organization’s capacity to leverage AI tomorrow.

Be prepared to “pick and stick”. Generative AI needs to be baked into your customer journeys. It needs to ingest established business rules and processes and operate on the contextual insights provided through an interrogation of available data sets.
You will fail, learn, iterate and improve.

Investing in the right partner and being realistic about their speed to competence will help set your expectations early and form the cornerstone of a productive relationship with your preferred AI partner.

Like any emerging technology and services industry, it can be difficult to know when your organization or your technology partners have reached an appropriate level of maturity to get into Generative AI. This can slow inertia and stifle innovation. It’s worth exploring non-traditional ways of contracting with your partners. For example, if your use case involves using generative AI for automating a feature, can you put a value on it and share the benefits? This type of arrangement reduces your organization’s exposure to cost, shares the risk of benefit realization and provides an opportunity to your preferred AI vendor partner to demonstrate their capability.

The best advice though is to be sure that before any of this – before you think about AI as a tool – your business must have a responsible AI policy framework that supports the security of customers’ data as much as it defends your own business rules. This is the foundation of customer trust when it comes to AI. You don’t want to be off your tricycle and under the Ferrari!