As artificial intelligence continues its rapid evolution, the language used to describe it is changing too. You may have noticed an increasing use of the term ‘AI agent’ as the integration of generative AI capabilities into business operations increases.
In short, AI agents are software programmes that use artificial intelligence to perform specific tasks autonomously, or with minimal human input. They can learn, adapt, and take actions independently rather than just responding to predefined commands.
Competition within the large language model (LLM) market has reduced the cost of creating AI agents, leading to the emergence of fast-growing startups. There was an AI boom in 2024 with startups building revenue at record pace, according to research by payments firm Stripe.
In their annual letter, Stripe’s CEO and co-founders Patrick and John Collison, wrote: “Much as SaaS started horizontal and then went vertical (first Salesforce and then Toast), we’re seeing a similar dynamic playing out in AI: we started with ChatGPT, but are now seeing a proliferation of industry-specific tools.”
“We see these new industry-specific AI tools as ensuring that individual industries can properly realise the economic impact of LLMs, and that the contextual, data, and workflow integration will prove enduringly valuable.”
The AI agent boom is a progression from vertical SaaS platforms which are used by 60 percent of small businesses in the US, wrote the Collison brothers.
Industry success stories: AI agents in action
AI agents are already used in many industries. In manufacturing, Siemens deploys AI agents for predictive maintenance which facilitates a conversation between the user, AI, and maintenance experts. The AI agent draws on a much wider pool of data than was previously possible to improve the decision-making process.
In banking, JP Morgan & Chase has developed various AI agents, one being ‘LOXM’ that executes high-frequency trades autonomously, reacting to market volatility faster than human traders.
In the hospitality industry, the promise of AI agents to perform personalised customer service at scale, workflow automations such as marketing campaigns, and various business intelligence applications, has, in many cases, been held back by legacy tech stacks and data silos. But hotel companies are now taking action to change this.
“A fragmented tech stack limits AI's ability to analyse cross-departmental data, reducing its capacity to enhance efficiency,” commented David Lodge, vice president engineering at IBS Software.
“Unifying CRS, PMS, RMS, booking engines, distribution channels, and loyalty programme data and operations through a single platform creates the foundation for AI to deliver transformative benefits,” he added.
Building and deploying an AI agent requires an ‘AI agent system’ that integrates generative AI, classical machine learning, and multiple data sources.
This integrated and open system allows AI agents to perform. So, companies need to lay the necessary groundwork to benefit from AI agents.
Wyndham Hotels & Resorts has worked with global CRM specialist Salesforce, to unify and eliminate the duplicates of 165 million guest records that were previously scattered across separate data sources. Last year, Salesforce integrated Wyndham’s reservations and loyalty systems and bridged the gap between individual properties and its corporate call centre.
Similarly, IHG Hotels & Resorts has standardised its CRM, loyalty programme, AI, and other data sources onto a single Salesforce platform.
With its unified platform in place, Wyndham was able to deploy an AI chatbot called ‘Einstein.’ Scott Strickland, Wyndham’s chief commercial officer, said: “Einstein has increased efficiency by generating recommendations on chat, drafting email responses, and delivering knowledge articles with a level of speed we haven’t seen before.”
The AI chatbot automatically analyses the context of customer requests plus account data and generates appropriate text responses. Human agents can edit or share the replies in one click, reducing average handling time by 90 seconds.
Unified systems pave the way for autonomous agents
In February 2025, Apaleo, the API-first open property management platform, launched an AI agent marketplace for the hospitality sector. The marketplace allows hospitality businesses to choose AI solutions from a variety of vendors for their specific needs and integrate them without costly system overhauls, Apaleo said in a statement.
German group mk hotels was overwhelmed with repetitive phone inquiries about reservations and check-in information. This reduced staff efficiency and led to missed booking opportunities, particularly outside business hours, as the front desk couldn’t manage calls 24/7.
The group integrated a voicebot from Plainview AI into its reservations system that responds to callers, checks availability and sends SMS confirmation links.
At first, most callers mistook the voicebot for a standard voice response system, leading to confusion and requests to speak to a human.
The hotel group refined the voicebot so that it clearly explained its abilities at the outset and responses were adjusted to feel more conversational and natural. The hotel reportedly saw a return on investment within one month and the voicebot was rolled out across multiple properties.
Other AI agents in Apaleo’s marketplace handle bookings via email; analyse sales data and recommend sales strategies; upload content into property management systems; and extract VIP guests and birthdays from a CRM database.
Other use cases for AI agents include software development; financial forecasting, modelling, and due diligence; fraud detection, employee training, and supply chain management.
In September 2024, Salesforce launched a service called Agentforce which enables businesses to create autonomous AI agents. These agents are equipped with trusted data, predefined actions, and guardrails to ensure they operate effectively and securely, the company said.
“The era of static SaaS is evolving into a dynamic, AI-powered future defined and accelerated by autonomous and agent-driven enterprises,” commented Jan Erik Aase, partner and global leader at research company ISG.
He added: “Adopting an agent-first model is a seismic shift for organisations, not just technologically but culturally. As autonomous agents become smarter, more efficient and more agile, entire customer interaction workflows will be redefined, and new decision-making processes will follow suit.”