How CSPs can take advantage of GenAI to supercharge diverse users’ needs
In recent years, we have observed the extraordinary evolution of artificial intelligence (AI) and its capacity to deliver remarkable business and operational advantages to organizations.
The most recent development is the emergence of generative AI (GenAI), which has risen to prominence as one of AI’s most compelling articulations.
One technology, endless possibilities
Generative AI is a subset of AI that focuses on creating data rather than just analyzing it, hence – ‘generative.’
To date it has demonstrated impressive capabilities for producing text, images, audio, and video content that closely resembles that which only human beings could create, and sometimes even better.
From writing code to making travel plans and awe-inspiring art, it’s no surprise that the popularity of platforms and tools such as ChatGPT, Bard, Scribe, Jasper, and others are experiencing explosive popularity.
And with its ability to deliver predictive models, automate tasks, and extract insights from unprecedented masses of data, media and communications service providers are also eager to embark on the exhilarating GenAI journey.
The user-GenAI engagement
Many different types of users across a CSP’s organization can gain significant benefits from generative AI. Applications that are driven by this technology can optimize users’ decision making, streamline their workflows, and enhance productivity.
Whether they are customer care agents, network engineers, financial analysts, or product managers, the potential of generative AI to supercharge skillsets across the service provider’s organization is profound.
But achieving its benefits and maximizing the value of GenAI requires selecting the right approach as to how users will engage with the technology. There are more ways than one.
In this article, I will cover the four approaches to engagement, what typifies each, and for which goals each is best suited, enabling CSPs to decide what’s best for them, when, and why.
- The embedded AI approach
This approach entails integrating generative AI seamlessly into existing applications, making it an integral part of the user’s workflow without disrupting the experience.
Users can access AI-generated insights and suggestions right from the application they’re already using today, engaging with existing interfaces. This eliminates the need for external tools or switching between platforms.
Examples include using buttons that create text that describes a graph and autocomplete suggestions.
- The AI-assisted approach
The AI-assisted approach, also known as “co-pilot,” serves to support users with an intelligent assistant. This GenAI co-pilot delivers valuable suggestions and insights to users, optimizing their decision-making and problem-solving.
The AI-assistant analyzes data, context, and user interactions. Through a panel that appears on the user’s app interface, it offers recommendations and timely and context-aware information. Moreover, it can even perform actions to further enhance efficiency.
For example, it can suggest relevant responses to customer care professionals as based on the conversation's real-time context, augmenting the care agent's ability to tailor resolutions, accelerating response times, and enhancing customer satisfaction.
- The AI-centric approach
In an AI-centric approach, applications are designed and developed from the ground up with generative AI capabilities.
Users rely heavily on AI-generated content and outputs for their tasks, and the application is optimized to make the most of AI-driven insights.
- The conversational approach
The conversational UI approach entails an interface that enables users to engage with generative AI through natural language conversations.
The interface is specifically designed to facilitate a human-like conversation between users and AI with the aim of completing a specific task with ease, speed, and efficiency.
To summarize, the four approaches to GenAI-user engagement as presented in an app’s interface are:
| Embedded AI | AI-assisted | AI-centric | Conversational |
|---|---|---|---|
| GenAI insights and suggestions are integrated into existing apps and workflows, e.g., buttons. | Co-pilot supports users with context-aware suggestions via panel, may perform tasks as well. | Applications are developed ground-up with GenAI capabilities and designed for easily accessing and acting upon insights. | The interface is designed for users to engage with GenAI through natural language conversations. |
How to decide which is best for the task?
When it comes to evaluating which approach will bring users and the organization the most value, the two parameters that can be most instructive are – interaction flow and content presentation.'
Interaction flow is about the user’s journey as they progress through the interface of the GenAI-powered app.
- A linear interaction flow is one that is structured and sequential, with a predefined, step-by-step path.
- A non-linear interaction flow delivers a dynamic and flexible user experience, enabling users to explore, navigate, and interact with the content or features in a non-sequential manner.
Content presentation refers to the way information is presented to users. It encompasses the design and layout of content which includes text and visual elements.
- Textual content presentation refers to a user interface that is driven by text, such as text-based instructions, labels, or contents.
- Visual content presentation refers to an emphasis on visual elements, such as images, icons, graphics, or visual cues.
How each approach measures up
Analyzing the four engagement approaches provides us with a deeper understanding of which is best suited for optimizing user experiences and outcomes across different applications and domains.
How each approach serves users, presents content and aligns with various interaction types:
| Evaluation parameter | Embedded AI | AI-assisted | AI-centric | Conversational |
|---|---|---|---|---|
|
|
|
|
|
Some tasks are better served by a linear/visual UI, while others by non-linear/textual, and so on. For example, since customer care agents handle a diverse range of inquiries, their GenAI engagement UI should include GenAI-powered dashboards that are non-linear, allowing for agility and flexibility. These dashboards should be able to present both textual content, such as support tickets, as well as visual elements, such as charts, for better issue management.
On the other hand, sales management apps should follow a linear flow to support the sales force with step-by-step guidance that is both textual, such as sales data, and visual insights, such as the sales funnel.
Ultimately, it depends on the service provider’s needs, objectives, and workflows, as well as how their users prefer to consume content. It is important to remember that in the same agent-facing application, it is common to encounter diverse sections that fall under different positions on the Interaction Flow and Content Presentation axes. Due to the multifaceted nature of these applications and the varying tasks agents perform, certain sections may exhibit a linear interaction flow, guiding agents through structured processes, while others may adopt a non-linear flow to accommodate more flexible and dynamic interactions. Similarly, content presentation within different sections may vary, as some sections may predominantly display textual information, such as customer records or support notes, while others emphasize visual elements, such as interactive charts and data visualizations for data analysis. This diversity in interaction flow and content presentation within the same app highlights the significance of implementing user-centered UI approaches tailored to each section's unique requirements, optimizing the user experience and efficiency across the entire agent-facing application.
In conclusion
The role of GenAI for optimizing business and operational performance is growing and will only continue to take a more prominent role. Yet, maximizing the value we can gain from it greatly relies not only on the power of the technology, but on how users engage with it.
This is why it’s so important to understand our users, their goals, and how they need to access and apply insights. Not every approach to engagement is right for every scenario.
Some users and tasks require step-by-step guidance, while others require flexibility. Some rely on text-driven insights, while others require diagrams. Certain industries may require specialized LLMs tailored to their unique needs.
The good news is – there’s an engagement approach for everyone. And the power of GenAI to drive excellence is only going to continue to reach even greater heights.
And personally, I can’t wait to see what’s waiting for us around the corner.
Amdocs Low-Code Experience Platform is a single platform serving all applications and channels, enabling service providers to configure their customer and agent journeys easily and flexibly in accordance with business needs.
Amdocs amAIz combines carrier-grade architecture leveraging open-source technology with large language AI models, creating a foundation for global communications service providers, enabling them to benefit from the immense potential of generative AI.
