2023 was a crazy year for AI. After years of watching this technology evolve from the sidelines, many of us have finally started to understand how it could impact our professional lives. It was a year full of triumphs and challenges, characterised by skyrocketing valuations for AI startups, widespread adoption by users and tech giants, and even a CEO drama that had us taking out the popcorn.

The integration of AI into various business sectors is reshaping many industries priorities. Among these, the Travel & Expense (T&E) sector slowly started to adopt modern AI technologies as it began to redefine the landscape of expense management.

It’s not that AI didn’t exist and wasn’t used before 2023. Several AI technologies such as Machine Learning and deep-learning have been enhancing numerous tools for quite some time. For us, this resulted in better OCR solutions to capture receipt data and helped build auditing solutions that could detect out of policy and even fraudulent spend much faster.

The difference now is that, in the last 24 months, we’ve been able to access newer technologies that have inspired us, and many players in the industry, to help build even better solutions that will become essential to get rid of the headaches around T&E.

The Evolution of Expense Reporting

Looking back, the manual process of expense reporting seems almost archaic. Employees painstakingly collected receipts, manually filled out forms, and endured lengthy approval processes. Finance teams used to live in a paper-based world, even earning the title of ‘paper pushers’ due to their extensive time spent on tedious manual tasks. This approach was not only time-consuming but fraught with risks of errors and fraud, necessitating exhaustive checks.

The advancements in technology brought a welcome change with digital spreadsheets, offering some organisation and efficiency. However, these still required manual data entry for each invoice or report, presenting a new set of challenges.

Today, systems like Rydoo’s use advanced OCR to efficiently process key invoice and receipt data, significantly reducing the need for human intervention.

The quest for efficiency led businesses to explore innovative technological solutions. In-house tools and Enterprise Resource Planning (ERP) systems with specialised expense management modules began to emerge, offering more integrated approaches.

The leap in cloud and mobile technology further transformed the landscape. Early adopters like Rydoo developed mobile apps for various operating systems, enabling on-the-go expense management — I still remember a time when we had our app in the Blackberry store! The introduction of Optical Character Recognition (OCR) initially faced limitations, but as AI and machine learning evolved, OCR’s accuracy improved dramatically. Today, systems like Rydoo’s use advanced OCR to efficiently process key invoice and receipt data, significantly reducing the need for human intervention.

And now, with integrated virtual and physical cards, a modern and efficient expense management solution is helping a lot to automate reconciliation, decrease the use of out-of-pocket spend and increase the level of control with pre-set limits and pre-spend approval. Thanks to the innovation in the field of open banking, software providers have been able to bring these payment solutions into a single platform.

So cloud, mobile, OCR and card technologies have been helping us innovate in these last years and will continue to do so. But it’s that AI will help us reach a frictionless way of managing our expenses even faster.

Envisioning the Future of T&E Management with AI

After working at the Rydoo for over nine years, I’ve had a backstage view of how challenging and time-consuming expense management can be for both the end-users and finance teams. Technological improvements in the last 30 to 40 years have made a big impact.

At Rydoo, we believe that a great T&E tool should feel like an assistant and a consultant for each employee. This means that it should help you manage every cent you spend as much as possible and provide insights to make better decisions.

AI systems will at some point in time leverage much more data from cards, receipts, calendars, travel itineraries, historical data,  etc, linking expenses to specific trips and projects, and auto-filling details. This way, they will only ask for your help when it’s actually needed. This level of automation not only saves time, but it can also enhance the accuracy of your organisation’s expense reporting.

In a few years, we should all be able to pay for a cup of coffee and do absolutely. The expense will be automatically processed, approved, reported and, if necessary, reimbursed on the same day.

Let’s take a closer look at some of the capabilities Artificial Intelligence can bring to the T&E business in the very near future, and how each can help make expense management smarter, faster and more efficient.

AI Autofill

If you’re working with modern smart expense software, there are some things you naturally expect from it. And as of today, with the latest advancements in technology, it’s safe to say that the manual entry of details is on the brink of becoming minimal, if not obsolete.

Last year, we introduced AI-Autofill as a feature preview at Rydoo. Next to the standard fields of date, amount and currency it also helps fill out the specific category, country and VAT of each expense. Unlike previous models that relied solely on the merchant category, leading to a high error rate, Rydoo’s AI autofill technology delves deeper, as it analyses the content of the receipt itself and uses your account’s history, to accurately fill in the category and country.

Previous models would put a Texaco receipt in the fuel category, while, in reality, the purchase could have also been food, drinks or even window washer fluid. With this new technology, Rydoo’s software can even detect if it was a late night dinner at the office or a lunch during a business trip. It will eventually help us auto-fill almost all the details of your expenses.

This approach will represent a significant leap forward, reducing the time and effort required for expense reporting, whilst also minimising the likelihood of errors.

Auto Approval and detecting out of policy spend

Why would we want a manager to review in-policy spend? Should we not rely more on pre-trip approval and post-trip insights? If you’re eligible to request reimbursements for paid parking when visiting a customer, it doesn’t always make economic sense to ask a manager to review and approve that receipt manually.

Today, we see many of our customers using auto-approval for these in-policy no-brainers, and AI will help us go even further. We can make sure all necessary checks are being done so the proof of payment is valid, that there is no conflict with the company’s policy and that there are no suspicious items on the receipt.

Manager approval will still make sense for out-of-policy spend or for high amounts that can have a significant impact on the budget. Of course, we want to make sure every manager gets the right insights frequently so they know their team is using common sense and using company resources appropriately.

Thanks to AI technology, we’ve been achieving great results in ensuring that employees comply with company policies. Some organisations have really complex regulations with a lot of details that should always be taken into consideration. But, when handling expenses manually, some things are bound to fall through the cracks, and using AI to analyse expense receipts and invoices can help with that. 

Imagine a scenario where you have a team member taking a train from Paris to Brussels. They buy a first class ticket and submit the expense, even though they’re aware that company policy doesn’t approve of first class trips. But the company is based in the US  and the ticket is in French. This is where AI comes in. The technology analyses the data from the ticket, understands that it’s first class and flags it to the end-user.

The same thing happens with detecting alcohol consumption or tobacco on a receipt. If an employee has a snack whilst on a trip and submits a ticket in a different language where a pack of cigarettes was listed, the technology can easily detect that. In just a few seconds, AI can identify what was bought and it summarises the total amount that was spent on tobacco on that meal.

This precision in categorisation and the ability to navigate complex company policies reflect AI’s expanding role. It also shows how the technology can help make the Finance teams’ jobs a lot easier, as they’re not obliged to know the names of all the drinks and brands available in other countries.

AI-Driven Fraud Detection in T&E

While auto-filling capabilities keep getting better and better, AI is also revolutionising the way organisations detect and prevent fraud in T&E expenses. The traditional methods of manual checking and random audits are giving way to sophisticated AI algorithms, which are capable of analysing vast amounts of data for irregularities.

These systems check spending patterns, authenticity of the receipt, merchant details and receipt line items, compare them with historical data, and flag anomalies that might indicate fraudulent activities. By identifying discrepancies and unusual patterns, AI-driven fraud detection tools provide an additional layer of security, safeguarding companies from potential financial losses and compliance risks. This proactive approach to fraud detection not only enhances financial integrity but also will prevent misuse of company resources.

AI-Powered Personalised Expense Management

Looking further ahead, AI is set to offer a more personalised and proactive T&E management experience. Imagine an AI assistant that not only tracks your expenses but also offers recommendations on cost-saving measures based on your spending habits and company policies. Can you take a taxi home after an after work drink? Just ask our policy assistant who has all the details of your policy so you don’t have to browse your SharePoint for that T&E policy and it can even check your company’s historical data to see if these kinds of expenses are usually rejected or require pre-approval.

Based on your calendar and travel bookings it could automate your reporting, linking each expense to the right trip or project. It should also give you and the whole organisation insights on how to optimise your team’s spending and ecological footprint.

Will AI automate everything?

All of these advances that we’ve been witnessing over the last few years with AI can be of great advantage in the world of T&E. Looking ahead, there’s still a lot of great innovation to come and a lot of unknowns. But overall we’re confident that AI can truly help organisations reach great results, save Finance teams a lot of time and even save on costs.

Leveraging AI is about entering a new era of operations and innovating core business operations, achieving higher levels of efficiency and strategic insights.

That’s not to say that it won’t have its challenges. Whilst AI has been making great progress, more ambiguous scenarios will still demand some human interaction. This balance might not always be easy to achieve, but it will become essential to maintain accuracy, compliance and to achieve the best possible outcomes for organisations.

Leveraging AI in the travel and expense management sphere is much more than just embracing technological advancements. It’s about entering a new era of operations all together and innovating core business operations, achieving higher levels of efficiency and strategic insights.

Moving forward, we can only expect AI to continue revolutionising our industry and we’ll make sure it can help us improve the way we work, and the way we live.