An expense report is a document (digital or paper) prepared by employees of a company to report expenses incurred in a work context. Expense report writing is a common process in many business settings, especially in larger companies. The following article describes the expense report writing process and its main steps. It focuses on the problems of manual processing and the benefits of automated expense report processing.

The expense report process

The expense report process is necessary whenever employees anticipate costs that are to be reimbursed by the company. It is a delicate and complex process that typically includes: settlement steps for actual and lump-sum expenses, supporting documents and more. Correctness and timing of reimbursement turn out to be of utmost importance, so it is critical to have a secure, reliable and digitized process to avoid delays and errors.

In general, the expense report is that process of ascertaining, verifying, and accounting for these anticipated costs and finally proceeding to reimbursement. It is an important process for companies and auditing agencies such as the Internal Revenue Service as it contributes to the formation of business income.

Examples of expenses typically covered by reimbursement follow:

• room and board expenses: catering and subsistence while away. For lump sum amounts, company policies should be checked so that non-reimbursable costs are not incurred.
• Travel expenses: for travel by plane, train, ship and so on. With one's own car, fuel and mileage reimbursement is used according to ACI tables.
• incidental or otherexpenses: not ascribable to the above such as parking, representation, photocopies of judgments filed by counsel, etc.

Stages of the expense note process

The expense report process is generically structured into the following steps:

• Drafting of the expense report: this step involves the creation of the document for the employee's claim for reimbursement to the principal. Drafting is the sole responsibility of the employee. Typically the compilation of the document requires a set of timely information that can typically be summarized as:
• Biographical and business data;
• List of expenses. Cost items incurred in travel with specific indication of amounts and any sequential number.
• Dates, time and place;
• Fiscal evidence such as: receipts, invoices, or travel tickets. The purpose is to prove actual economic outlay for expenses related to travel;
• Useful comments and notes to justify extraordinary costs.
• Supervision: the supervision phase consists of checking the accuracy of the information entered for reimbursement. This phase can be carried out through internal company resources or outsourced. The auditor, in addition to verifying whether the information entered is correct and whether there is a receipt with fiscal validity, confirms whether the expenses to be reimbursed are in compliance with company regulations and policies.
• Administration: the last step is to transfer the validated expense report data to the company's accounting system. This is used to make the accounting entry and subsequently, to reimburse the employee.

To be able to properly manage the expense report process, it is critical to be able to nimbly process the multitude of different purchase receipts that may be involved in this process.

Processing of purchase receipts

This section examines the types of documents pertaining to the expense report process and the major complications that have to date limited the automation of the expense report process in different contexts.

Documents involved

In the expense report management process, the documents involved as receipts are many and often quite different from each other. They are typically semi-structured documents that often contain some of the necessary information useful in compiling the expense report itself. For proper optimization of the process, it is essential to be able to classify and obtain the useful information automatically. Below are the most commonly involved document types: receipts, invoices, parking tickets, highway tolls etc.

‍

‍

Main difficulties

Each of the above documents is a semi-structured document that depending on the vendor can also change a great deal in terms of the location of the information of interest and the format. For a large company, another problem related to centralized expense report management is inherent to language: in fact, purchase receipts are typically related to travel to different countries with the associated change of language, currency and format. In some cases, receipts are even handwritten which further complicates processing (both manual and automated). Finally, receipts are often delivered in different formats such as scans (often degraded), pdfs, and photographs from cell phones (commonly difficult to read).

Main stages

Class identification

One of the relevant pieces of information to identify when filling out the expense report is precisely the category related to the receipt that is being attached to the expense report. Typically, the classes considered eligible fall into the following list:

• Entertainment expenses (hairdresser, clothes etc...);
• Transportation (plane, train, cab, vehicle rental);
• Overnight stays;
• Breakfast/Lunch/Business Dinner;
• Travel tolls;
• Parking;
• Expenses related to conferences/meetings and events.

This information is often not explicitly in the document but can be inferred by reading and identifying the class to which it belongs (e.g., for an overnight stay, one will check that the receipt is actually from a hotel).

Relevant information to be extracted

Much other relevant information can be directly deduced from the tax slips being, many times, contained within them. The information that is commonly extracted for both accounting/tax purposes and for control purposes is as follows: seller information, seller's full name, vat/vat number, country and address information; currency of payment, total, taxes paid, mode of payment, any purchase lines, and finally the date time and place.

How to process purchase receipts

Manual approach

Extracting data from purchase receipts can be costly, time-consuming and error-prone if done manually. The processing steps require well-trained people who can identify relevant information in the invoice and extract it consistently from sometimes complex layouts. Some challenges and issues related to manual processing include:

1. Cost Issues
   Although manual extraction of data from documents may work for small companies with limited processing piers, it becomes expensive as the company expands. In many cases, manual data extraction involves hidden costs in addition to those associated with hiring more employees to perform data extraction. Adding more employees to the workflow often leads to coordination problems, which in turn can lead to errors, especially in identifying and entering data. Data validation itself is a critical step that adds to the cost component of processing. In fact, extracting information without verification steps can have error rates as high as 4 percent. The 1-10-100 data entry rule is well known in data entry back offices: verifying data accuracy at the point of entry costs about $1, cleaning up errors by rechecking the entire batch of data costs $10, and escaped errors cost the company $100 or more.
2. Time Issues
   Manual data extraction can be time-consuming, especially global supply chains, as multiple checks and approvals are required and teams in different countries often need to be coordinated. Integrating the various processing and verification steps can also become complicated because of the various people involved in the hierarchy and the approval levels required.
3. Human fallibility
   The data entry process is repetitive and tedious and can be demoralizing. In addition, purchase receipts do not have a standard format. Although much of the information listed in the previous sections is present in all documents, each vendor uses a different format with significant spatial variability in the data. Manual data extraction therefore can be laborious given these variabilities. Language variations between the place of issue and place of delivery can also be a significant challenge in understanding the data.

Traditional OCR solutions

The processing of purchase receipts using traditional OCR techniques and template matching/regex is a decidedly ill-advised and wasteful approach as it is necessary to have ad hoc rule sets and templates for each document type. The formats are many and the vendors potentially number undefined in advance. The languages to be considered are often numerous for a solution that must work in processes with global reach. This makes the number of rules or templates needed definitely numerous if not infinite and constantly changing as new formats and countries are considered. All this results in a high cost of setup and maintenance of the solution and also often poor performance. In addition, maintenance and configuration of the solution must be done by trained resources with technical training.

Intelligent Document Processing (IDP)

A modern approach based on Deep Learning techniques is the best choice for solving such problems. The ability to use the best techniques of Computer Vision for image analysis and NLP for natural language understanding allows scaling to a high volume of different formats and languages without having to adapt the solution each time (writing new rules or configuring new templates) but simply having a sufficient amount of data to instruct the system. Such an approach can also benefit strongly from the human validation step, which in addition to correcting errors made by the system can enable continuous learning of the algorithm allowing it to improve over time and calibrate itself to the specific process.

Compared to traditional solutions, even adding a new field that you want to extract, adding a document category to classify, or wanting to add a new language among those supported does not involve writing code. The collection of new documents will be sufficient, and the subsequent retraining of the system can also be followed by a resource without technical skills but who has knowledge of the process. Finally, the most effective IDP solutions also enable unprecedented accuracy of results that far exceeds both manual processing and traditional solutions.

myBiros and benefits

myBiros is an Intelligent Document Processing solution that enables automatic document processing. Core functionalities are information extraction and automatic document classification. All of this is offered through a prebuilt set of ready-to-use APIs with pre-trained templates for common use cases and the ability to retrain the entire pipeline (both the OCR engine and the document interpretation system) for custom cases.

By leveraging advanced deep learning techniques that analyze multimodal features, it is possible to process all document types with a single solution. The system uses pre-trained models, data-augmentation techniques, and for that reason can be trained with a small volume of data allowing even processes involving a small volume of documents to be automated.

This solution includes a scoring mechanism: the system reduces false positives by enabling the ability to review low confidence data while minimizing errors. Interaction with a human user enables the system to correct errors while continuing to train the system so that past mistakes are not repeated(human in the loop and continuous learning). Finally, the high scalability of the cloud-based architecture allows highly variable masses of documents to be processed without having to allocate expensive resources in advance.

The features mentioned so far allow myBiros to perform optimally on the documents involved in expense report processing: receipts, invoices, parking tickets, highway tolls. If you are curious about how myBiros works in order to simplify the processing of any purchase receipt, contact us and try our demo. We are ready to help you!

‍