Attend almost any hospitality technology conference at the moment and you will hear a familiar promise: artificial intelligence will finally tie everything together.
It’s an attractive idea. New connectors allow AI to access information from property management systems, building controls, utility meters and finance platforms. Emerging standards such as Model Context Protocol (MCP) are making it easier for AI to access different sources of data and tools, reason across them and respond in plain language. For hotel operators who have spent years wrestling with systems that refuse to talk to one another, this is a significant step forward.
But there is an important distinction between using AI to interpret integrated data and relying on AI to be the integration itself. MCP isn’t the problem. It is a valuable way of making data and tools accessible to AI. The risk comes when AI is simply “thrown in” and expected to decide which systems to consult, whether it has all the information it needs and what conclusions should be drawn.
At Cadae, we’re firmly in favour of AI – we’re an AI-first business ourselves.But we believe it works best when it sits within a structured system designed to make its outputs dependable.For hotels, that matters because operational decisions ultimately rely on trust. Whether you’re looking at an efficiency benchmark, investment recommendation or ESG report, you need confidence that the information behind it is reliable.
There are three areas in particular where structure matters: consistency, completeness and correction.
With a conventional integration, the same information processed in the same way should produce the same result. That predictability matters when hotels are reporting performance, demonstrating compliance or making investment decisions.
AI models don’t always behave in the same way. Ask the same question twice and an AI may interpret the available information differently or produce a slightly different response. Updates to the underlying model can also change how it behaves over time.
For an everyday query, that variation may not matter. But if an answer is informing a major investment decision, an ESG disclosure or a report to hotel ownership, it becomes far more important.
Hotels need to be able to understand where their figures came from and reproduce them when necessary. If an auditor or owner asks how a calculation was reached, the answer needs to stand up to scrutiny.
That requires AI to operate within a system with defined rules and processes, rather than being left to determine the methodology each time.
The second challenge is completeness. AI can only work with the information available to it. The difficulty is that it may not always recognise when an important piece of the picture is missing.
An experienced hotel engineer might immediately question energy data from a spa hotel that shows no consumption associated with its leisure facilities.Their experience tells them there should be a pool plant somewhere.
Similarly, imagine a hotel wing was closed for refurbishment for three months, laundry was outsourced halfway through the year or part of the property’s energy consumption wasn’t being metered.
Without that context, an AI system could still produce a perfectly plausible benchmark. The problem is that it would be benchmarking an incomplete picture.This is one of the biggest challenges with AI: a convincing answer isn’t necessarily a complete one.
The role of a well-designed system is therefore not simply to give AI more data, but to understand what data should be present and flag when something doesn’t look right.
The third issue is what happens after a recommendation has been made.If a hotel invests in an energy-saving measure, the obvious next question is: did it actually work?
A credible performance system should compare predictions with real-world outcomes. If predicted savings don’t materialise, it should investigate why. If they exceed expectations, that should also inform future decisions.
An AI agent operating independently doesn’t automatically create that feedback loop. It can reach a conclusion, deliver a confident recommendation and move on without ever discovering whether it was right. For hotels, that is a missed opportunity.
The real value comes when AI operates within a system that continually compares expectations with actual performance. That’s how confidence improves over time – through evidence, not simply because an AI model sounds certain.
This is why the debate shouldn’t be framed as AI versus traditional integration, or MCP versus another approach.
AI and technologies such as MCP can make it dramatically easier to access, connect and interpret information across a hotel. But they become much more powerful when they operate within a structured system that provides the rules, context and feedback needed to make their outputs dependable.
That’s the approach we’ve taken at Cadae.Rather than simply producing a confident-looking score, Cadae considers the completeness of the information behind an assessment. Missing or unusual data can be identified and flagged instead of being quietly absorbed into a result that may not reflect the full picture.
As actual consumption data flows back into the platform, predictions can also be compared with what happened in reality. Over time, the system’s understanding of an individual property becomes stronger and confidence grows because conclusions have been tested, rather than simply asserted.
For hotel operators, that’s the distinction that matters. AI shouldn’t simply be added as the latest technology panacea and expected to solve years of fragmented systems overnight. Equally, hotels shouldn’t be afraid of embracing it. Used within the right structure, AI has enormous potential to help operators understand complex data, identify opportunities and make better decisions.
The hotels that get the most from AI won’t necessarily be those that adopt it fastest. They’ll be the ones that build the right foundations underneath it.
For more information, visit Cadae.
Attend almost any hospitality technology conference at the moment and you will hear a familiar promise: artificial intelligence will finally tie everything together.
It’s an attractive idea. New connectors allow AI to access information from property management systems, building controls, utility meters and finance platforms. Emerging standards such as Model Context Protocol (MCP) are making it easier for AI to access different sources of data and tools, reason across them and respond in plain language. For hotel operators who have spent years wrestling with systems that refuse to talk to one another, this is a significant step forward.
But there is an important distinction between using AI to interpret integrated data and relying on AI to be the integration itself. MCP isn’t the problem. It is a valuable way of making data and tools accessible to AI. The risk comes when AI is simply “thrown in” and expected to decide which systems to consult, whether it has all the information it needs and what conclusions should be drawn.
At Cadae, we’re firmly in favour of AI – we’re an AI-first business ourselves.But we believe it works best when it sits within a structured system designed to make its outputs dependable.For hotels, that matters because operational decisions ultimately rely on trust. Whether you’re looking at an efficiency benchmark, investment recommendation or ESG report, you need confidence that the information behind it is reliable.
There are three areas in particular where structure matters: consistency, completeness and correction.
With a conventional integration, the same information processed in the same way should produce the same result. That predictability matters when hotels are reporting performance, demonstrating compliance or making investment decisions.
AI models don’t always behave in the same way. Ask the same question twice and an AI may interpret the available information differently or produce a slightly different response. Updates to the underlying model can also change how it behaves over time.
For an everyday query, that variation may not matter. But if an answer is informing a major investment decision, an ESG disclosure or a report to hotel ownership, it becomes far more important.
Hotels need to be able to understand where their figures came from and reproduce them when necessary. If an auditor or owner asks how a calculation was reached, the answer needs to stand up to scrutiny.
That requires AI to operate within a system with defined rules and processes, rather than being left to determine the methodology each time.
The second challenge is completeness. AI can only work with the information available to it. The difficulty is that it may not always recognise when an important piece of the picture is missing.
An experienced hotel engineer might immediately question energy data from a spa hotel that shows no consumption associated with its leisure facilities.Their experience tells them there should be a pool plant somewhere.
Similarly, imagine a hotel wing was closed for refurbishment for three months, laundry was outsourced halfway through the year or part of the property’s energy consumption wasn’t being metered.
Without that context, an AI system could still produce a perfectly plausible benchmark. The problem is that it would be benchmarking an incomplete picture.This is one of the biggest challenges with AI: a convincing answer isn’t necessarily a complete one.
The role of a well-designed system is therefore not simply to give AI more data, but to understand what data should be present and flag when something doesn’t look right.
The third issue is what happens after a recommendation has been made.If a hotel invests in an energy-saving measure, the obvious next question is: did it actually work?
A credible performance system should compare predictions with real-world outcomes. If predicted savings don’t materialise, it should investigate why. If they exceed expectations, that should also inform future decisions.
An AI agent operating independently doesn’t automatically create that feedback loop. It can reach a conclusion, deliver a confident recommendation and move on without ever discovering whether it was right. For hotels, that is a missed opportunity.
The real value comes when AI operates within a system that continually compares expectations with actual performance. That’s how confidence improves over time – through evidence, not simply because an AI model sounds certain.
This is why the debate shouldn’t be framed as AI versus traditional integration, or MCP versus another approach.
AI and technologies such as MCP can make it dramatically easier to access, connect and interpret information across a hotel. But they become much more powerful when they operate within a structured system that provides the rules, context and feedback needed to make their outputs dependable.
That’s the approach we’ve taken at Cadae.Rather than simply producing a confident-looking score, Cadae considers the completeness of the information behind an assessment. Missing or unusual data can be identified and flagged instead of being quietly absorbed into a result that may not reflect the full picture.
As actual consumption data flows back into the platform, predictions can also be compared with what happened in reality. Over time, the system’s understanding of an individual property becomes stronger and confidence grows because conclusions have been tested, rather than simply asserted.
For hotel operators, that’s the distinction that matters. AI shouldn’t simply be added as the latest technology panacea and expected to solve years of fragmented systems overnight. Equally, hotels shouldn’t be afraid of embracing it. Used within the right structure, AI has enormous potential to help operators understand complex data, identify opportunities and make better decisions.
The hotels that get the most from AI won’t necessarily be those that adopt it fastest. They’ll be the ones that build the right foundations underneath it.
For more information, visit Cadae.
It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout. The point of using Lorem Ipsum is that it has a more-or-less normal distribution of letters, as opposed to using ‘Content here, content here’, making it look like readable English. Many desktop publishing packages and web page editors now use Lorem Ipsum as their default model text, and a search for ‘lorem ipsum’ will uncover many web sites still in their infancy.
It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout. The point of using Lorem Ipsum is that it has a more-or-less normal distribution of letters, as opposed to using ‘Content here, content here’, making it look like readable English. Many desktop publishing packages and web page editors now use Lorem Ipsum as their default model text, and a search for ‘lorem ipsum’ will uncover many web sites still in their infancy.
The point of using Lorem Ipsum is that it has a more-or-less normal distribution of letters, as opposed to using ‘Content here, content here’, making
The point of using Lorem Ipsum is that it has a more-or-less normal distribution of letters, as opposed to using ‘Content here, content here’, making it look like readable English. Many desktop publishing packages and web page editors now use Lorem Ipsum as their default model text, and a search for ‘lorem ipsum’ will uncover many web sites still in their infancy.
It is a long established fact that a reader will be distracted by the readable content of a page when looking at its layout. The point of using Lorem Ipsum is that it has a more-or-less normal distribution
Copyright BlazeThemes. 2023
