What you need to know

• Major Australian firms, including Telstra, ANZ and KPMG, are leveraging generative AI technologies to enhance business operations and customer interactions.
• Telstra's AI tools, Ask Telstra and One Sentence Summary, have improved the effectiveness of 90% of customer service agents in trials, reducing follow-up calls by 20% and aiding in faster onboarding for new staff.
• Following the launch earlier this year of its AI Immersion Centre, ANZ Bank is training 3000 leaders in generative AI at its Melbourne headquarters, aiming to integrate AI across all business lines for improved customer service and operational efficiency.
• KPMG has developed the Comply AI platform to streamline ESG compliance assessments, significantly reducing the time needed for manual evaluations and enhancing collaboration among experts.
• A common theme across all the implementations is they are using Generative AI tools to improve workflow efficiency, allowing employees to focus on more complex tasks, thereby increasing job satisfaction and overall productivity.
• All three companies are planning broader implementations of generative AI, anticipating further enhancements in operational capabilities and the ability to meet regulatory demands, especially with the upcoming mandatory ESG reporting requirements.
• The results can be found in a long list of customer case studies shared online by a senior Microsoft marketing exec.

Australia's corporate giants are harnessing the transformative power of generative AI, with Telstra, ANZ and KPMG reporting solid early wins that enhance customer service, streamline operations and redefine compliance. All three want to expand the role of generative AI across the enterprise.

According to Telstra, generative AI tools called Ask Telstra and One Sentence Summary have boosted the effectiveness of 90 per cent of agents in trial involving several hundred staff, and helped slash the number of follow up calls needed to resolve customer queries.  . 

Rohit Lakhotia, General Manager of Customer and Channel AI at Telstra: “One of our benchmarks is solving the problem on the first call. Our customer service agents are on the frontlines. As the first point of contact with Telstra, they’re the face of the company, so it's crucial that we support them with the right tools to have the most successful customer interactions possible.” 

The details are confirmed in online documents shared by Microsoft’s Alysa Taylor,  Chief Marketing Officer, Commercial Cloud & AI recently outlining the use of co-pilots among the software giants top customers.

At Telstra, the complexity of sourcing answers from a vast pool of information can be a time-consuming process, particularly for new customer service agents. Built on Microsoft’s Azure infrastructure, Ask Telstra tool has been fine-tuned to access Telstra’s extensive internal knowledge base, providing customer service agents with quick access to necessary data. In mid-2023, approximately 200 customer service agents participated in a trial of the solution, and their feedback was instrumental in refining the tool before a broader rollout across call centres and retail locations.

“New agents found that Ask Telstra gave them a big boost to becoming more effective, faster,” said Lakhotia. “When we gave them the Ask Telstra tool, they spent less time asking for help from their managers, who frequently were busy with other conversations. It’s been very helpful in onboarding new staff.”

Similarly, One Sentence Summary, tested by 100 agents in 2023, provides quick access to a concise summary of recent customer interactions. This tool appears as a clickable button within the customer’s history in Telstra’s CRM software. Agents can click the One Sentence Summary button to reveal a clean interface that succinctly outlines the customer’s situation, thereby reducing the time needed to understand customer needs and facilitating better transitions between agents.

Another features of One Sentence Summary is its ability to deliver sensitive customer information efficiently. Agents have reported an enhanced capacity to engage with customers experiencing challenging circumstances, such as those needing financial assistance.

Full deployment

Although the full deployment of Ask Telstra and One Sentence Summary is planned for early 2025, pilot programs have already begun to enhance customer experiences and streamline agent workflows. “Ninety per cent of customer service agents who tested One Sentence Summary increased their effectiveness,” said Lakhotia. “Their calls required 20 per cent less follow-up than those handled without the tool. Over 80 per cent of the agents trialing Ask Telstra agreed it had a positive impact on customer interactions.”

The company’s goal is to propagate AI more widely throughout the organisation. “We’re unlocking value and growth as technologies like generative AI proliferate at speed,” added Kim Krogh Andersen, Group Executive of Product and Technology at Telstra.

Meanwhile, Orla Glynn, Telstra Executive Data & AI Transformation, Acceleration & Adoption, said the goal of building an AI-fuelled organisation "is very much a journey of continuous learning”. 

“The pace of change over the past year has been truly remarkable. It’s crucial that we adapt, collaborate through partnerships, and share knowledge so that we can unlock the huge transformational potential offered by AI," Glynn said. “Looking ahead, as we advance in developing our own internal AI capabilities, we foresee an even greater opportunity to share our experience to scale and enhance the AI capabilities of our Telstra Enterprise and Telstra Business customers.”

ANZ Bank's AI Immersion Centre at its Melbourne headquarters aims to facilitate the understanding and safe application of generative AI among the bank’s leadership, with plans to train 3,000 leaders over the next year.

The initiative is designed to accelerate AI adoption within the bank, providing hands-on learning experiences that illustrate the potential of generative AI. To further support this initiative, ANZ has invested in 3,000 licenses for Copilot for Microsoft 365.

Gerard Florian, ANZ Group Executive of Technology, highlighted the bank's commitment to leveraging generative AI across its operations. “We are actively embracing generative AI across all lines of our business to improve how we serve and protect our customers, process work and meet regulatory standards. Applied safely, generative AI can enhance experiences for customers and employees, and help us generate business outcomes faster, with lower risk and at lower cost.”

According to Microsoft and LinkedIn’s 2024 Work Trend Index, Australia and New Zealand are leading the way in generative AI adoption in the workplace, with 84 per cent of employees utilising the technology, surpassing the global average of 75 per cent. Notably, many employees are independently integrating AI tools into their work environments, with 78 per cent in Australia and 81 per cent in New Zealand using their own AI solutions.

The AI Immersion Centre is designed to empower senior leaders at ANZ to foster innovation and support their teams in the adoption of AI tools, thereby transforming banking services. This training initiative reflects the bank's broader strategy to fully embrace an AI-enabled future.

The bank has already made significant investments in various generative AI technologies, including Copilot for 365, GitHub Copilot, and Copilot in Edge, to enhance productivity and innovation across its workforce.

Following a successful Copilot pilot in November 2023 - just as ChatGPT was being unleashed into the world, ANZ expanded its use by purchasing an additional 3,000 licenses. 

“Our strategic investment in Copilot for 365 is evidence of our commitment to harnessing the power of generative AI to enhance our services, streamline operations and empower our employees," Florian said.

Furthermore, ANZ has deployed GitHub Copilot to 3,000 software developers and engineers, realising productivity improvements as engineers complete programming tasks 40 per cent to 55 per cent faster. As the adoption of GitHub Copilot scales, ANZ anticipates significant increases in code production alongside improved overall code quality.

Additionally, the bank developed a private chatbot named Z-GPT, enabling employees to generate and integrate internal information. This tool, created by ANZ’s AI Centre of Experimentation in India played a crucial role in the bank's experimentation with generative AI using internal data subsets.

According to Florian, as the bank is embracing generative AI at speed and scale, security alongside innovation and technological advancement will be a key consideration.

The launch of the AI Immersion Centre in August this year followed the successful trials of Copilot for  365 and GitHub Copilot, as well as the development of Z-GPT.

KPMG Leverages AI to enhance compliance and efficiency

KPMG Australia deployed AI technologies to streamline various data discovery tasks for a decade, and is now expanding its generative AI investment. As an example, the consulting giant has developed a proof-of-concept for its ESG Comply AI platform.

Guided by natural language prompts input by subject matter experts (SMEs), the compliance checker executes advanced text analysis on numerous client source documents. This analysis determines full or partial compliance, or noncompliance, significantly reducing the time needed for manual assessments. KPMG aims to use Comply AI and Copilot to reformat employee workflows by accelerating data discovery, automating tedious administrative tasks, and fostering enhanced collaboration with clients and colleagues.

Throughout its AI journey, KPMG said it has prioritised the establishment of a trusted framework of policies and practices to ensure the responsible and ethical use of AI. Early on, it formed AI councils at both country and global levels to assess potential use cases against responsible AI criteria. These councils primarily evaluate whether solutions enhance productivity and efficiency, improve quality and accuracy, support employee and client wellbeing, and provide deeper insights.

Building on the value derived from early AI technologies like natural language processing and machine learning, KPMG member firms, led by KPMG Australia, opted to expand their inquiry into generative AI to expedite numerous manual tasks.

“We’ve created a trusted AI framework, used globally across KPMG, that considers the safety, risk, ethics, and other factors of venturing into generative AI to make sure that we’re doing it the right way and creating a safe environment for our staff and our clients,” said Sophie Papworth, Senior Manager of Research, Strategic Design, and Product Strategy at KPMG Australia.

KPMG tested generative AI technologies from various vendors, including participating in an early access program for Copilot for Microsoft 365. During this initial testing phase, the firm identified Copilot's potential for enhancing employee productivity and delivering high-quality outcomes. It subsequently decided to utilise Copilot, Azure OpenAI Service, and Azure AI Search as the foundation for many of its generative AI applications.

“We’ve done our research in terms of what generative AI technology we want to use,” said Shahab Dayani, Director of Data Analytics & AI at KPMG Australia. “We were able to sit down with senior leaders at Microsoft, discuss the opportunity, and learn about the latest and greatest generative AI capabilities that Microsoft has to offer, including Copilot.”

After hands-on experience with the products, KPMG found the transparency of the code in Copilot facilitated easier development and made the functionality more understandable for SMEs.

“In Copilot, you can actually understand what the AI tool is doing, rather than simply relying on results published to the user,” said Dayani. “We strongly believe that the AI capability that Microsoft, especially in the generative AI space, has to offer is first class.”

In early 2023, KPMG Australia initiated its journey into generative AI with the KymChat conversational AI assistant, followed by KymTax, which focuses on tax precedent. These efforts were soon complemented by the AI Workbench, a platform designed to host numerous responsible AI-powered tools aimed at improving individual productivity across KPMG's global workforce.

One of the first tools being developed within the AI Workbench is the Environment, Social, and Governance (ESG) compliance reporting assessment tool known as Comply AI. ESG reporting is crucial for customers, investors and financial institutions to understand the comprehensive impact of an organisation’s ESG practices. Historically voluntary, ESG reporting is set to become mandatory in many jurisdictions starting in 2024.

The impending mandatory ESG reporting requirements will significantly increase the workload for audit staff. To assist clients in meeting these demands efficiently and cost-effectively, KPMG created a pilot version of Comply AI to automate the initial ESG compliance assessment process under the guidance of KPMG SMEs.

“That doesn’t mean Comply AI is going to replace the task that was being done by an expert,” said Dayani. “It just makes experts a lot more efficient and gives them the capabilities to perform those tasks in a much quicker and higher-quality way for our clients.”

KPMG developed the proof-of-concept for Comply AI within a matter of weeks, then rolled out the solution to a select group of SMEs in its audit and assurance practice in late 2023 for testing against manual assessments.

“Pilot testing has enabled us to identify which features of KymChat and Copilot add the greatest value and shape a secure and multifaceted change program as we continue its rollout,” said Jason Fogaty, Partner, Connected Technology Group at KPMG Australia.

The use of Comply AI involves SMEs uploading client source documents in various formats, such as Word, PDF, and Excel, into the tool. A message extension in Microsoft Teams, powered by Azure OpenAI, conducts advanced text analysis on these documents. This analysis is then automatically compared to a comprehensive database of ESG standards, quickly producing an assessment that indicates whether documents are fully compliant, partially compliant, or out of compliance. The tool also provides direct links within the report to standards highlighting areas of noncompliance.

Upon completion of the multi-page assessment, SMEs can engage Copilot using natural language prompts to generate a detailed analysis summary of the entire report. This allows for seamless review of results with colleagues and clients via Teams chat, with the ability to drill down into specific standards for recommendations and discussions on next steps. Additionally, AI can be used to draft an email summarising the findings to peers or clients, along with to-do lists and follow-up meeting schedules.

“We’ve got all of these different components, but where Copilot really brings it all together is report generation and drafting emails to the client, which becomes really, really powerful for our staff,” said Dayani.

KPMG said it's using Copilot under strictly controlled conditions. The firm curated datasets behind the scenes to ensure more accurate output. The internal communications team employs these tools for drafting plans and materials, while the internal strategy team uses it to convert workshop materials into preliminary strategy drafts. Firm leaders have noted significant time savings when using Copilot for email responses.

The time savings are also evident in Comply AI testing, which has led to a notable reduction in the duration required for ESG assessments. The use of Copilot to summarise assessment results and automatically generate emails and action items for colleagues and clients accelerates the overall process.

“If you were spending hours and hours to summarise hundreds of pages of an assessment, now you can use KymChat and Copilot to do that in minutes,” said Dayani.

KPMG said its subject matter experts can now redirect the time saved from mundane tasks to focus on areas of partial compliance or noncompliance, creating broader insights that further mitigate client risk and drive success. This increase in efficiency enables KPMG to handle more business with its existing team, laying the groundwork for revenue growth.

By reallocating time saved from manual tasks to complex judgment-driven areas, the firm enhances service speed and quality for clients. KPMG also claimed the shift contributes to greater career satisfaction for employees, allowing them to engage in more meaningful aspects of their work.

“Our vision at KPMG, using generative AI and Copilot, is to enable our people to do some of the tasks they’re currently doing manually, in a much more efficient and productive way with higher quality,” said Dayani.

He insisted this approach does not replace the SME’s role; rather, it enhances it. “We see Copilot and the Comply AI platform as an enabler to help our experts more quickly review the output, validate it, and make sure that they agree with the results.”

The interoperability between Comply AI and Copilot within Teams enables real-time collaboration among SMEs, clients, and coworkers without requiring extensive training, as many clients and most KPMG personnel are already accustomed to Teams. This collaborative environment not only expedites processes but also enhances transparency, consistency, and service excellence.

KPMG envisioned the potential for Copilot and the Comply AI platform extending beyond immediate ESG reporting needs.

“Our plan is definitely to go broader than ESG compliance checking,” said Dayani. “We would like to bring in more use cases such as some that are impacting our financial services clients. Our ultimate vision is to make Comply AI and Copilot globally available so all of the other KPMG member firms - over tens of thousands of KPMG individuals and their clients - can benefit from it.”