With the democratization of AI, the value of data has never been more treasured. What are the opportunities, trends and challenges for businesses in Asia Pacific?
Businesses understand that new technologies, when implemented and utilized the right way, have the tremendous potential to drive revenue growth, achieve cost-savings, and unveil untapped opportunities. The increasing technology investments made by companies are a testament to this understanding.
Among these technologies, AI continues to capture significant attention. Notably, one aspect of AI that has been making headlines and dominating discussions is generative AI. Indeed, research has shown that generative AI has the potential to contribute as much as US$4 trillion to the global economy. Moreover, productivity improvements through generative AI use cases could amount to nearly US$8 trillion.
The remarkable figures and projections surrounding generative AI make it evident that this technology is poised to revolutionize the business landscape and drive massive value in Asia Pacific and globally. And generative AI thrives on data.
DigiconAsia discussed the trends and developments in data and AI with Steve McMillan, CEO, Teradata.
Where do you see the biggest opportunities in data and AI in the Asia Pacific market?
McMillan: The APJ region offers distinct opportunities in each market. In Australia and New Zealand, major financial institutions are aggressively advancing their journey to the cloud and looking at AI. Conversely, in Japan, organizations in some business fields are often adopting a hybrid environment, even as they actively look to utilize AI.
In the South Asia market, our strategy has focused on accelerating our strategic partnerships with organizations. Our partner-focused approach has been critical in bolstering our presence and business in this region. Today, we have a modern partner ecosystem to extend our reach and drive business outcomes for our customers, exemplified by our recent strategic partnership with Dell.
In a broader sense, we are observing considerable traction across the Asia Pacific and Japan region, particularly across cloud-native, multi-cloud, hybrid cloud and AI. Flexibility and customer choice have always been paramount at Teradata and this versatility aligns perfectly with the preferences of our customers across the region.
You mentioned Teradata’s open and connected approach to data management. Can you explain how this approach benefits organizations in Asia Pacific, and what makes it different from other data platforms in the market?
McMillan: Having an open and connected approach means that we encourage our customers to bring their preferred data models and tools. We then provide the platform to facilitate seamless connections between various data, tools and even clouds. In other words, data scientists are able to harness the power of data in combination with their preferred tools.
At Teradata, we believe that by leveraging as much clean, managed and governed data as possible, customers are able to retrieve trusted insights out of their data, regardless of its location in the ecosystem. We support cloud service providers (CSPs) like AWS, Azure and Google Cloud, but we can also do ecosystems that use both multi-cloud and hybrid cloud. This is especially important in APJ, where private on-premises clouds may be the ideal deployment method based on CSP availability.
This open and connected approach is unique to Teradata’s data platform and gives customers ultimate choice and flexibility in their ecosystem. Regardless of what they choose, Teradata’s analytics offering, ClearScape Analytics, is designed to offer open and connected analytics, where businesses can push down queries and run analytics at the (data) source to deliver real-time insights without the need to move data around the ecosystem.
What is your perspective on the role of generative AI, and how is the company leveraging generative AI to benefit enterprises?
McMillan: From our perspective, generative AI represents a pivotal step forward in AI innovation. This cutting-edge technology has the potential to fundamentally reshape how businesses operate. Across various sectors, including customer experience, marketing, sales, software engineering, and R&D, we are witnessing profound transformations.
Generative AI is unlocking the full potential of AI in ways that were previously unimaginable, compelling companies to consider its integration.
Our recent survey, conducted in collaboration with IDC, revealed that more than half of the 900 respondents feel a ‘high’ or ‘significant’ pressure to integrate generative AI within their organizations in the next 6 to 12 months. However, only 30% feel adequately prepared to leverage generative AI today, indicating a significant gap that needs to be bridged.
With years of expertise in the AI domain, Teradata stands as the go-to platform for AI implementation, providing the most cost-effective solution, proven performance and flexibility to innovate faster, enrich customer experiences, and deliver value. Our technologies not only facilitate the AI journey for businesses but also accelerate their realization of value. In fact, we are already helping our customers deploy AI solutions for supply chain management and improving customer experiences.
We have also embedded AI into our VantageCloud Lake product, which can speed the development of code for technical roles and deliver data-based insights to approved, non-technical roles that don’t know how to code, expanding the use of data and analytics throughout the organization.
With the widespread deployment of AI across enterprises, concerns regarding control, oversight, and intellectual property protection arise. How does Teradata help organizations in the region ensure accountability, security, and trust in their AI implementations?
McMillan: We are staunch proponents of the limitless possibilities AI and machine learning offer when wielded responsibly. Whether it is detecting fraud, optimizing manufacturing performance, ensuring preventive maintenance, or addressing a myriad of other challenges, the impact of AI and ML on business outcomes knows no bounds. But we recognize that the success of AI and ML projects depends on businesses having the right tools and procedures in place.
-
Enabling AI-driven automation and decision making: Accountability, compliance, and good stewardship are all required to deliver trusted AI that can positively impact customers and empower organizations. This is only possible with stellar data management and governance tools. 92% of companies surveyed said that data ethics and the responsible use of data is paramount, and 9 in 10 companies shared that they have a formal ethical data resource or board in place.
ClearScape Analytics offers tools for both governance and adherence to regulatory standards for organizations. With ClearScape Analytics, customers gain the ability to govern models at every stage of their lifecycle. This can be particularly challenging for organizations that deal with a vast number of models, such as per-segment modeling. It becomes crucial for businesses to detect any model drift, signaling when any of the numerous models deviate from the expected results and require adjustments. Our ModelOps capability plays a crucial role in ensuring that these models operate as anticipated, thereby fostering trust and reliability in their performance.
An open and connected platform is also critical for enabling trusted AI. Especially with a technology changing as quickly as AI has been, organizations need the flexibility to integrate with the most advanced services in the ecosystem, such as Microsoft Open AI Services. Lock-in doesn’t work when technology advancements are being made every single day and a platform that isn’t open to these changes won’t serve its customers very well in the long term.
-
Removing bias from AI: AI systems can exhibit biases that stem from their programming and data sources. Bias in machine learning creates a vicious cycle of discrimination caused by incorrect conclusions drawn from inaccurate predictions.
Part of the solution to reducing AI bias is to also ensure that the teams building the models are diverse and very involved in the model management process. Even in trusted AI models, humans are still the responsible party. Additionally, techniques such as using unbiased synthetic data, which can supersede real life examples, should be considered. It is also crucial that these teams maintain efforts throughout the lifespan of the model to monitor for bias that can happen via model drift.
-
Implementing trust in the reliability by building transparency and explainability into AI solutions: Transparency in AI/ML context refers to the comprehensibility of how these systems function. Simply put, it ensures that people can grasp the reasoning behind a model’s decisions. At Teradata, we strongly advocate for integrating transparency and explainability into AI solutions. We believe this integration is pivotal in establishing trust in the accuracy and fairness of a model’s predictions.
Our ModelOps service is tailored to assist companies in incorporating transparency and explainability into their AI/ML solutions. This support enables businesses to gain insights into the decision-making process of their models. By understanding how these decisions are reached, companies can have confidence in the reliability and fairness of their model’s predictions.
Given that banking and finance are major industries for Teradata in the Asia Pacific region, could you share examples of how Teradata has helped financial institutions in the region to innovate and advance their digital transformation or AI strategies?
McMillan: Banks in Asia Pacific work similarly to banks globally. Through our conversations with decision makers in the banking industry here, we found that many of them are reviewing their technology strategy.
Teradata has played a pivotal role in assisting financial businesses to implement their AI/ML models effectively. One financial service provider in Australia previously struggled with its one-pipeline-per-process approach for the deployment of their AI/ML models which resulted in a mere 30 predictive models being deployed over the span of a decade.
However, Teradata’s ClearScape Analytics, featuring its Bring Your Own Model (BYOM) capability, is changing the game. With this advanced solution, the Australian financial service provider is poised to deploy around 200 models in one go. These models are prioritized, ensuring rapid deployment on Teradata’s VantageCloud and seamless operationalization at scale. This shift made for a significant leap, moving from three models per year to a substantial 200 models at once.
This strategic move allows the financial service provider to thoroughly analyze their customer base and enhance their marketing campaigns. Vantage stands as a vital infrastructure in this customer’s AI/ML strategy, serving as the core platform for these transformative initiatives.
