- Citi is partnering with fintech company HighRadius to automate a key money-collection process for businesses.
- The companies have launched Citi Smart Match, which applies artificial intelligence and machine learning to the process of matching open invoices to received payments – painstaking and costly task known as “cash application.”
- The new feature enhances Citi’s crucial and highly lucrative Treasury and Trade Solutions (TTS) business, which provides daily financing and cash management services thousands of multinational corporations in nearly 100 countries.
Citigroup is deploying robots to help massive corporations collect payments from customers, the latest enhancement to the bank’s bread-and-butter business of managing the nitty gritty tasks for companies’ daily cash and transactions.
Citi on Thursday announced a partnership with Houston-based financial technology company HighRadius to roll out a new feature called Citi Smart Match that applies artificial intelligence and machine learning to the process of matching open invoices to received payments – automating and streamlining the otherwise painstaking and costly task.
It also further bolsters Citi’s crucial and highly lucrative Treasury and Trade Solutions (TTS) business, which provides daily financing and cash management services to thousands of multinational corporations in nearly 100 countries and pulled in $8.5 billion in revenues in 2017.
When businesses receive payments, they have to verify which customer the cash came came from and match the incoming cash to the correct invoice or account so they know that customer is all paid up, a process known as “cash application.” It’s perhaps obvious how important it is to get this right, as a businesses’ operations could quickly grind to a halt if it doesn’t know with high accuracy which customers are paying their bills. It would also waste a lot of time and energy trying to collect money from customers who already paid – potentially upsetting a loyal customer in the process.
It’s also gotten a lot more complicated with the proliferation of electronic payments, as money can come in from a variety of places over the Internet separate from the invoice – or a customer could send in a lump sum payment that covers multiple invoices. So cash application is often done manually by specialists within accounts receivable teams to ensure accuracy.
This is where Citi and HighRadius come in.
Citi is the day-to-day banker for many the world’s largest corporations across the globe, handling an array of recurring money and payment needs. Last year, the company put out a call to the fintech community to come up with solutions for the clunkiness of cash application.
“We’re in the business of providing new, creative solutions to our customers,” Manish Kohli, global head of payments and receivables in Citi’s Treasury and Trade Solutions division, told Business Insider. “The problem of cash application is not a new problem – it’s been a problem for quite some time.”
Citi quickly zeroed in on HighRadius, an enterprise software firm focused on applying AI and machine learning to the payment receivables business.
In February, Citi’s venture-capital arm invested in HighRadius, and over the past six months the companies collaborated to build Citi Smart Match, which promises to automate cash application for Citi’s TTS customers.
Here are some of nuts and bolts of how Citi Smart Match works and what it will be able to do:
- Use AI to capture data from checks: “The technology is able to learn where to look for information and how to identify it.”
- Identify and trove through emails and email attachments for key details: “The AI learns how to identify emails containing remittance information by examining keywords and attachments. Similar to paper processing, the system can identify and extract the data fields relevant to reconciling accounts. It can do this in situations where the remittance is provided in the body of the email as well as when it is in an attached file.”
- Organize the automated incoming stream of data: “The adoption of EDI, or Electronic Data Interchange, which can allow for full automation of account reconciliation from participating customers, specifies a number of formats for everything from invoices to debit authorizations. This will offer client a broad range of ways to organize incoming data.”
- Build custom rules for cleaning up and formatting data so invoices are recognized and matched automatically: “Once the data is extracted, it is necessary to cleanse it so that it is easily recognizable by Client’s accounting system and all the invoices are automatically matched and cleared without the analyst having to perform any manual activities. The formatting and cleansing of data can be done using a sequence of rules which transform the extracted raw data into a format that matches open invoices and can be automatically processed by the ERP.”
After all this, the machine-learning technology “will learn from patterns and behaviors of payments received and continually improve the matching rates over time,” according to the companies.
While this may not be happy news for cash application specialists who might find their services less prized, it’s good news for large businesses, which could potentially save both time and millions in costs, which could be invested elsewhere.
“It is significant. There’s the potential for 60, 70, 80% of their cash application costs going away thanks to automation,” Kohli said. The match rate for payments to invoices would continue to improve over time, too, thanks to the machine learning capabilities, Kohli added.
Kohli noted that manpower savings wasn’t the only benefit for Citi’s customers: It helps customers reconcile invoices faster and reduce the average collection period, thus enabling them to free up credit lines for their own customers more quickly. And it also reduces the error rate that typcially comes alongside manual matching.
Citi and HighRadius are currently piloting Citi Smart Match with some early adopters and expect to roll out the service more broadly in the coming weeks.
This post has been updated from its original version.