Uploading bills, checking the data, filling in the account codes and approving supplier bills, are all the repetitive tasks that Xero users need to do on a routine basis. These tasks are time consuming and ongoing, a major pain point for accounting. But they are also perfect candidates for machine learning, which is a software system that uses data to adjust decision making rules.
Michael Fitzmaurice, the co-founder and chief developer of EzzyBills technology, is determined to make your job easy. With Artificial Intelligence and machine learning, EzzyBills has captured the imagination of many Xero users and accounting service providers and is putting your Xero accounts on auto-pilot…
“I came across Xero when my family moved back to Australia after working in Silicon Valley for ten years.
When consulting for businesses, I realised that small businesses desperately need technology to improve their business efficiency. Yet, they often lacked the funds to pay for customised software.
The Xero ecosystem makes it possible for software companies to develop cutting-edge software for small businesses collectively, therefore individual business only pays a small fraction of the development cost.”
EzzyBills made a big leap in technology with a system designed to process invoices like a person. It has been learning how to do so from over two million invoices and continues to learn each
day. Therefore, EzzyBills asks no questions from a new user before processing your first batch of invoices. Machine Learning is the core of the EzzyBills invoice processing technology.
Discover an Invoice Processing system that is simple to use
The use of Machine Learning makes EzzyBills stand out among the crowded Xero add-on marketplace. The software is fast and accurate, having no rival in terms of the ability to extract line items with description. This is precisely why many accounting professionals are loving it.
The simple user interface adds to the beauty of EzzyBills technology. It does not have long lists of
user input and it does not need customisation to read line details. It takes 3 minutes to sign up a new account and process your first couple of invoices. “The truth is: the more a software does for you, the less you need to do.” said Michael from EzzyBills.
“New Xero users discover and adopt EzzyBills every day. They nd it easy to understand and appreciate the software as the software “thinks” in the same way as they do. As John Spensor, from ESMSL service, a Commercial and Domestic Plumbers and Heating Engineers based in Middlesbrough, put it, EzzyBills is ‘A program with good service and common sense.’ When this level of simplicity is achieved by our software, we know we have done a great job.”
Nearly all the invoices that EzzyBills process come automatically via email. Users forward their bills or sale invoices to their unique EzzyBills email address. Invoices may come in as an attachment, an in-body invoice or a hyperlink. If it is some spam or advertising material, they will be discarded. If it is a statement, order confirmation and remittance advices, they go through a different workflow – typically get stored and not processed.
The invoice data is then extracted, and optionally approved by a project or office manager. Xero account codes and tracking categories are determined from your invoice history, and if you are using Purchase Orders, EzzyBills will automatically mark them as billed.
New Feature Release – EzzyBills Review
At this point of reading, you may want to say, “If EzzyBills does everything, what do I need to do?” Well, you should review a small proportion of invoices that EzzyBills has singled out for you.
EzzyBills is smart enough to know when it is 100% confident, or when it is not. For example, when we have processed many bills from the same supplier and they were all correct in the past (by cross validating against the payed invoices), then EzzyBills is 100% sure that the invoice is extracted correctly and there is no need for you to check.
However, some other conditions may lead EzzyBills to believe that an invoice will need review. Examples of these conditions are:
• the bill is for a new supplier,
• the system is not confident about some of the extracted data,
• previous invoices from this supplier had their account codes corrected.
So instead of reviewing all the invoices processed, you only need to review a few. See an example
in Figure 1. With machine learning, the number of invoices that need to be reviewed will drop in time, eventually dropping to nearly none.
Another great feature of EzzyBills is ‘Price Check’. If the supplier price has changed, EzzyBills is able to detect it and notify you. EzzyBills also supports a multiple-level approval process.
Businesses that have a large number of invoices to process will benefit greatly from using EzzyBills. They are generally in service industry, such as retails, restaurants, hotels, trades, pharmacies, and medical practices.