Based on GlobalData’s conversations with technology vendors, service providers and enterprises, customer experience (from customer service to personalisation) has been the top business driver for the use of AI. There are already many AI-enabled use cases including chatbots and agent assistants. With GenAI developing rapidly, there is a general expectation that AI will improve experiences through better understanding of questions and customer demand. In reality, however, many AI projects and pilots have not moved into production (as high as 85%, according to some technology vendors). There are many reasons for this, ranging from cost to data quality, governance and security.
GlobalData expects the development of agentic AI to be a key driver in turning large language models into revenue generators. Agentic AI offers benefits associated with problem-solving, not just content generation. It is able to operate autonomously, making decisions and executing tasks. AI agents will continue to learn and self-optimise, thereby enhancing its effectiveness over time.
Technology vendors are now introducing tools to help enterprises build agents without extensive development work. One example is Salesforce with the launch of Agentforce in 2024. The company has reported growing market traction with Agentforce, and at its recent event held in Sydney, Australia, the company shared customer examples and the business benefits the solution has delivered.
Service providers have a key role as advisors to enterprises
One example is Fisher & Paykel, a global household appliances company that has implemented Agentforce to enhance its customer service. By giving digital agents access to the company’s knowledge base of more than 10,000 articles, these agents have been able to handle more customer inquiries (increasing from 33% of inquiries being handled by chatbots which had previously been introduced up to 60% with Agentforce). Using Agentforce, Fisher & Paykel’s digital agents are now able to address clients’ appliance issues with access to customer information for context, engage customers through conversation and schedule technician visits without fully walking through the issue.
Hipages is another featured Salesforce customer. The company connects tradespeople with homeowners for home improvement through an online platform. Hipages uses Data Cloud and Agentforce to reduce turnaround time for onboarding verification checks from about three hours to near real-time. The verification process is necessary to ensure tradesmen have the necessary certifications required under various regulations. Replacing manual processes with AI has enabled Hipages to scale its operations, provide tradesmen with faster turnaround times and reduce human error, all without increasing its workforce significantly.
It is not only Salesforce innovating in this area. Other technology vendors are also establishing platforms to help enterprises build agents with minimal amount of effort (such as through a low-code platform). In December 2024, Google Cloud launched Agentspace, which offers a low-code visual tool to build expert agents that can assist those in business roles, such as business analysts, HR professionals, software engineers and marketers, to become more efficient. Agentspace leverages Google’s multimodal AI search capabilities, NotebookLM (an AI-powered research and note-taking tool), and pre-built connectors such as Confluence, Google Drive, Jira, Microsoft SharePoint and ServiceNow. Microsoft has also introduced Azure AI Agent Service which brings together models, data, tools and services to deploy AI agents.
While technology vendors are making efforts to enable ease of deployment of agents, the challenge for enterprises going forward will be the choice of technology partners and to keep up with the pace of innovation. Service providers should play a key role as advisors to enterprises, supporting them in developing strategy for implementation and ongoing support. It is crucial for service providers to have the necessary expertise and certifications across different vendor environments since it is likely that businesses will use multiple vendors as their operations evolve.