Cambridge Service Alliance (CSA) hosted its Industry Day last week. As a partner of the research organization, HCLTech attended along with 35 senior leaders from some of the UK’s leading enterprises.

Throughout the day, four CSA researchers presented their insights on how AI and generative AI (GenAI) are impacting the future of service in business.

1. AI-driven personalization in service, Jan Blümel

Blümel highlighted that today’s forward-looking organizations are focusing on applying AI to generate more humanistic communication and identify emotions in digital customer service interactions. The goal is to create a personal touch in customer service, starting with text-based interactions and expanding to voice and multimodal language models.

The key aspects he covered included:

• Understanding the customer's previous experiences and emotions to adapt responses effectively. This involves analyzing factors like frustration levels and recurring issues.
• Integrating empathy and personalization while balancing efficiency metrics like resolution time. The research explores how to incrementally introduce AI capabilities to maintain a personal touch.
• Using conversational analytics to identify elements like empathy and emotions, and then applying conversational coaching to improve responses based on customer profiles.
• Leveraging previous interactions to personalize responses, including using speech characteristics like pitch and articulation to understand customer profiles.
• Exploring the potential of multimodal language models that can combine voice and text data to enhance the personal touch in customer service interactions.
   

2. Moving toward a CX-centric strategy, Gautam Jha

In his presentation, Jha delved into the challenges of managing customer experience (CX) and the need for a more holistic, CX-centric strategy within organizations. He highlighted that the traditional, siloed and interaction-focused approach to CX is no longer sufficient, as businesses are investing heavily in CX initiatives but not always seeing the expected returns. 

The core issue, according to Jha, is that managing CX requires a multifaceted approach, involving the alignment of the organization's CX strategy, vision, design elements and the integration of data and emerging technologies like AI. Stakeholder alignment is also vital and various roles within the organization need to be brought together to solve this complex puzzle. 

To address these challenges, Jha proposed a research framework with three key building blocks: attitudes (shifting mindsets from product-centric to customer-centric), capabilities (integrating siloed data and skills into “cohesive customer journey value streams”) and methods (optimizing digital tactics, service design and data science through a CX lens). 

Underpinning this framework is the theory, “The Mundanity of Excellence,” which suggests that small, often overlooked details can make a significant difference in achieving excellence.

3. Optimizing service experience design with generative AI, Helen Zhao

In the early stages of her research project, Zhao’s presentation discussed the potential of GenAI in transforming the way organizations approach service design and customer experience innovation. Well known applications of GenAI are the ability to generate content like text, audio, synthetic data and imagery, essentially providing “intelligence as a service” to organizations. 

Examples of how GenAI can be applied include text summarization, idea brainstorming, content creation and code generation. Zhao highlighted that these capabilities have the potential to automate up to 70% of employee work activities today, transforming how organizations approach knowledge-based tasks. 

“It's very different from previous AI technologies, where it's less about focusing on repetitive work, more about knowledge work, and how it's being used to transform industries,” she said.

Looking ahead, Zhao highlighted the need to develop a coherent framework that integrates this technology seamlessly into the design thinking process, while addressing these critical challenges.

4. Enhancing customer service in manufacturing with LLMs and RAG systems, Abhimanyu Kanwar

The final presentation focused on using AI and machine learning to enhance customer service in the manufacturing industry. Kanwar shared insights from his research on developing an AI-powered “co-pilot” system to assist field service engineers and reduce customer service response times. 

The key elements include defining personas (operators, technicians and supervisors), building a comprehensive knowledge base leveraging technical documentation and historical ticketing data and utilizing large language models (LLMs) and retrieval-augmented generation (RAG) techniques, which improves generative language models by enabling access to relevant information from external and internal data sources to deliver more precise and contextually appropriate responses.

Initial pilot testing of the system has shown promising results, according to Kanwar, with an 80% correct response rate for sample support tickets. 

However, he did acknowledge the trade-off between cost-effectiveness and efficiency, as more sophisticated, knowledge graph-based systems that automate most prompts may offer greater optimization potential but require significantly more investment in preparation. 

The aim of the research is to identify whether this route with a huge initial investment or a cheaper, base model with more labor costs is the best system for the future.

Reflections: Immense potential but integration challenges

Among main takeaways from the four research presentations, it was evident while the AI and GenAI hold immense potential in transforming service, the true challenge lies in integrating it seamlessly into the organization's culture, mindsets and ways of working. By focusing on the smaller details, aligning stakeholders and keeping the human at the center, businesses can unlock the full power of AI and GenAI to deliver exceptional customer and user experiences, while driving long-term, sustainable growth.