ai:sight > Annual Volume 2023 > GenAI in action, from call centers to healthcare

GenAI in action, from call centers to healthcare

Generative AI’s ability to automatically generate new text, image, audio and video content is jumpstarting innovations across multiple industries.

You’ve probably heard the buzz surrounding Generative AI (GenAI) lately. It’s been the talk of the town, with media coverage ranging from awe-inspiring tales of its limitless potential to concerns about its impact on society. But let’s dive deeper into this groundbreaking technology, swiftly making its presence felt across various industries.

Think about call centers, for instance. Thanks to GenAI, chatbots have taken a giant leap forward. These intelligent bots can now provide detailed answers to your questions and even learn from your responses. It’s like having a virtual assistant constantly evolving to serve your needs better – and who is conversing with you like a human. That’s what we call convenience!

Yet, GenAI continues. Creative marketing and media professionals have also harnessed its power to generate captivating content across different mediums. From eye-catching text to mesmerizing images, stunning soundtracks to mind-blowing special effects, GenAI is revolutionizing the way professionals create. Some even use it as a springboard for their creative process, starting with an automatically generated first draft that humans can refine and perfect. Humans are very good at improving things once they are given something to start with. It’s like having an AI collaborator that sparks your imagination by giving you the first draft!

However, the software development sector may have witnessed the most significant disruption caused by GenAI. Here is an interesting fact:

When the Italian Government recently banned ChatGPT, the productivity of the country’s software developers plummeted by 50% in just two days. That’s the power of GenAI in action!

It has become an integral tool for developers, enhancing efficiency and speeding up the creation of innovative software solutions.

GenAI is a true game-changer. “It creates a level playing field that enables people less familiar with data science or AI to converse naturally with their data and make sense of it,” he explained. “Imagine effortlessly interacting with complex data sets, regardless of your background or technical expertise. It’s like having a personal data whisperer by your side.

Ajoy Singh

Chief AI Officer, Fractal

Nonetheless, GenAI is at its nascency, and organizations worldwide are eager to explore its incredible potential. This groundbreaking technology can engage in hyper-personalized conversations with customers, unlocking precious wisdom buried within vast amounts of complex and unstructured data. But its impact continues beyond this.

GenAI’s lightning-fast search capabilities across thousands of documents are poised to accelerate research in healthcare and pharmaceuticals revolutionizing how we uncover vital insights.

Himanshu Nautiyal, Chief Product Officer at Fractal, shares, “Many clients today seek voice-based conversations with their data. They want to ask questions and receive narrated answers, complete with specific points and highlighted actions. This capability empowers them to make informed decisions and focus their efforts more effectively. Instead of passively receiving information, they want to engage in meaningful dialogue with their data, driving their discoveries.”

As the saying goes, “With great power comes great responsibility.” This couldn’t be truer regarding the disruptive force of GenAI. We’re accustomed to working with software that assists humans in collaboration, but now, machines actively participate in conversation for the first time. This transformative shift will undoubtedly reshape how we work and our jobs. However, it presents a significant challenge for organizations adopting this technology. They must ensure the resulting disruption is positive rather than traumatic for those involved.

To address this challenge, two fundamental principles hold the key to success:

  • First, do no harm.
  • Design with people, for people.

When dealing with a disruptive phenomenon like GenAI, prioritizing the principle of doing no harm becomes the foundational pillar for all subsequent capabilities and efficiency gains. This principle has been exemplified through Fractal’s work with Fortune 500 clients spanning various industries. The equation is simple: to mitigate risk, every action taken must be defensible, responsible, transparent, and unbiased. Coupled with an AI, engineering, and design approach that places people at the heart of decision-making, this forms the bedrock of any successful AI implementation.

“When technology is something that’s done to people, it feels like it’s going to take their job. But if it’s something that’s done with and for them, it feels like a superpower,” said Ann Hintzman, Chief Design Officer at Fractal. “We have a huge opportunity to create solutions that are emotionally resonant with how people want to work and the purpose they find in their work. This starts with people, what they are trying to achieve, what they value, and how they want to grow. That means designing with those people, starting with an ethical standpoint and using technology to augment that.

One question for those adopting GenAI today is how to ensure transparency and explainability in a technology designed to provide human-like responses.

“One of the key challenges with GenAI is that its models will provide varying responses that are not fully explainable,” said one of our clients from a global investment firm. “The way to solve that is to tame it so that we know how to handle exceptions. That means doing much coding to ensure the GenAI model has the right prompting. How we enable the parameters will be the difference between a great AI product and a dumb one.”

“We can readily adopt GenAI for internal use, but the moment we start exposing it to real business practices, legal, compliance, and information security factors all come into play. For example, suppose we enable our website with GenAI so financial advisors and direct customers can discover our products more easily. How do we ensure the different disclosure document classes are appropriate for those users? How do we identify and protect customers who mistakenly type in personally identifiable information? Some of those challenges will be significant, and they all need careful thought before we open a two-way, GenAI-enabled conversation with clients.”

But with appropriate parameters in place, GenAI has great potential to help start conversations with customers and boost staff efficiency.

GenAI can help to create simpler, more dynamic marketing content for our clients. Instead of presenting clients with fact sheets or product documents, we could enable a more interactive discussion with chatbots. It can be as simple as replacing a static ‘good morning’ on the website with a different, AI-generated greeting each time they visit. These interactions could change the customer experience because more general conversations make product discovery much easier. Internally, GenAI can also help our people to be more efficient. It can help marketing, sales, and distribution teams to find product-specific information, for example. When investment managers track a particular stock, they can get a lot more detail in a very conversational manner. Then the next step is to determine how the technology can be used to make decisions and research to help those people outperform.

Direct Line Group, one of the UK’s largest insurers, is looking at how GenAI can help to achieve its vision of a world where insurance is personal, inclusive, and a force for good. Enhancing the customer experience is an ideal starting point.

There are use cases for GenAI across the entire insurance lifecycle. GenAI can help make our customer journeys quicker and easier, for example, by offering context-sensitive, personalized help as customers fill out information. It can also help summarize and simplify the language used in the policy documentation. Little things like these add up to making all our lives easier. Further down the line, when something unfortunate happens and a claim is made, GenAI can help us speed up the path to getting our customers back on the road or back in their homes, for instance, by automating time-consuming processes and distilling documents to extract high-value data. We are exploring all these areas with the guiding principle of making things better for customers and employees.

Jonathan Saunders

Chief Data Officer at Direct Line Group

Insurance and technology are continually evolving, and as the industry finds new uses for GenAI, staying on top of the risks is essential. 

Like humans, AI models can get things wrong,” Saunders said. “In parallel to exploring the exciting opportunities the technology brings, we all need to be aware of the risks and have mitigations to ensure that AI models always do the right thing. At Direct Line, we have robust data and AI ethics frameworks and a sophisticated control framework to ensure privacy by design and constant model monitoring. Like every other organization, we must constantly update policies and processes as the technology develops to ensure it is held to the highest standards. GenAI consumes a lot of data and carries out much processing, so it’s very important to think widely about the risks. For example, we must look at the environmental impact and assess the value holistically, all the way through the technology supply chain.

Companies planning to use GenAI have a unique opportunity to help shape its evolution. By collaborating with other business and academic institutions, Direct Line is actively developing the technology and the ethics surrounding it. 

We work with many companies and universities at the cutting edge of AI,” Saunders said. “Through collaboration and shared excitement about the opportunities we now see, we stay on top of the state of the art. We’re also very active in the cross-industry shaping of data and AI ethics. It’s incumbent on all organizations to treat both the technology and the accompanying ethical implications with equal weight – one should not exist without the other.”

For those seeking to embark on the GenAI journey, it’s essential to view the business through two distinct lenses that provide valuable insights and guide decision-making:

  • Risk Assessment: Consider the potential risks when implementing GenAI. An in-depth analysis of your organization will aid in distinguishing between core and non-core aspects. Presently, many organizations concentrate on automating non-core or internal tasks, such as document processing and data entry, to address unforeseen risks and gaps in the existing frameworks. This approach significantly reduces human effort and minimizes associated risks automatically.

    However, in sectors like banking, core GenAI business activities involve underwriting, pricing, and selling loans, which are only partially automated. It is crucial to evaluate factors such as reputation, data privacy, and regulatory compliance upfront and to understand the associated risks thoroughly. Doing so will enable the implementation of proactive measures to mitigate these risks effectively.
  • Incremental impact: Explore the positive incremental impact that GenAI can have on your business. Identify areas where it can enhance operations, productivity, and customer experiences. This lens helps prioritize use cases and solution areas for initial implementation.

Ajoy shared, “The significance of these lenses is that they help guide organizations towards suitable starting points. The most appropriate use cases and solution areas can be identified by carefully examining the business through these perspectives. Let’s take the example of a bank: implementing GenAI for anti-money laundering checks proves to be a low-risk, high-impact area, making it an ideal starting point. On the other hand, underwriting carries a higher impact potential, but it also comes with greater risks, such as the possibility of being accused of unfair lending practices. A more cautious and experimental approach may be warranted.”

To harness the true potential of GenAI, it’s crucial to delve into the underlying processes of each use case. Doing so gives a clearer picture of how GenAI can make a difference. Take, for instance, the fact that large language models excel at deciphering unstructured data rather than structured, numeric data. This explains why their capabilities shine brightest in creative tasks and environments such as call centers, where a quick understanding of emotional context drives both quality and productivity.

Suraj Amonkar, Client Partner and Head of AI@scale, Machine Vision, and Conversational AI at Fractal, acknowledges this space’s complexity and rapid evolution. “There are myriad possibilities with GenAI, but it requires fine-tuning through pretraining and specific utilization,” Amonkar explains. “Understanding how to overcome the challenges associated with these complex technologies by reducing friction and enabling seamless integration for businesses is important.”

Fractal leads the charge in GenAI solutions, with offerings like Fractal GPT for seamless AI chatbot integration, Flyfish as the world’s first GenAI-powered sales assistant, and Crux Co-Pilot for dynamic voice-based business intelligence. Fractal also offers Genesis, a GenAI platform that can help enterprises implement and scale GenAI led transformations securely.

While the full extent of GenAI’s impact is yet to be realized, organizations must start preparing for the future. Innovation, driven by these transformative technologies, is set to accelerate. A steadfast commitment to ethical, people-centered design ensures that these developments genuinely benefit individuals and businesses.

Himanshu emphasizes the importance of readiness for the GenAI future. “AI can seem dehumanizing if not implemented thoughtfully. Hence, the approach should always be grounded in a human implementation, for humans.”

Understanding the underlying processes involved in the use case is essential to understand how GenAI can help. Today, for example, large language models are better at making sense of unstructured data than structured, numeric data. This is why those capabilities shine brightest in creative tasks and environments like the call center, where a quick understanding of emotional context drives quality and productivity. Further work on incorporating contextual data, intelligence, and workflows over the LLMs, has resulted in platforms like Crux Co-Pilot and Senseforth, which can drive value for enterprise users in safe, reliable, predictable ways by reimagining processes end-to-end, using both structured and unstructured data.

Rate this Article


Related Reads