- Key insights: BMO is exploring new use cases for AI. We’re using AI to help sales reps have more informed conversations with mid-market business customers.
- Expert quote: “We want to be a trusted advisor to our customers, and this allows us to fulfill that role in a more consistent way,” said Rose Grande, head of North American corporate card products and programs at Montreal-based BMO.
- Future outlook: BMO plans to begin a pilot in Canada in the coming months, with a full-scale rollout by the end of 2026.
BMO was looking to expand its business relationships in the middle market. This requires a detailed understanding of your customer’s business.
Processing details
Such insights aren’t always easy to come by, says Rose Grande, director of North American corporate card products and programs at Montreal-based BMO.
“We can do customer intelligence work in-house on a very bespoke, one-off basis, but as far as being able to scale that, we haven’t been able to do that,” Grande told American Banker.
She and her team decided to try Codat’s technology, which can capture payment and accounts payable information for a bank’s customers through an application programming interface and use machine learning to generate recommendations for BMO sales representatives.
BMO and Codat’s collaboration comes at a time when most U.S. banks are considering AI use cases and deploying AI pilots into production. in
“Banks certainly want to leverage AI to not just automate internal operations, but to better understand customer behavior, identify unmet needs, and make bank relationship managers more effective,” Bradley Leimer, founder and principal at Leimer One Advisors, told American Banker. “Commercial banking is a natural place for this because so much value resides in cash flow, payments, working capital, and customer-supplier relationships. Much of that data is fragmented in enterprise resource planning systems, accounting platforms, PDFs, spreadsheets, bankers’ meeting notes with customers in CRMs, etc. I’ve seen this myself in several use cases I worked on at Sumitomo Mitsui Banking Corporation. The data is hard to work with, but the rewards are big.”
Codat technology features
Codat helps BMO capture and analyze customer data that is difficult to obtain on its own.
“Getting customers to share their data is an inherent pain point across the banking industry,” Joey Raoult, chief revenue officer at Codat, told American Banker. “Bankers, financial managers, financial consultants, payment consultants, and sales professionals are constantly asking for data from their customers in order to better serve them.”
Codat captures details about a customer’s accounts payable, including which suppliers they are paying and how they are paying them. “All of these types of insights help us advise customers on payment best practices,” Grande said.
Codat automates data ingestion through direct API connections developed to retrieve information from more than 20 enterprise resource planning and accounting software programs, including QuickBooks, Oracle, NetSuite, Sage, Microsoft Dynamics, Workday, and Xero. Collecting client data using APIs is much more effective than asking customers to send CSV or PDF files, Rault says. To collect data from accounting and ERP software for which it has not built an API, Codat uses an “intelligent upload tool” that bank customers can use to upload raw files from those systems.
The vendor uses machine learning to classify transaction files and generate personalized recommendations that are shared with BMO’s treasury and payments sales teams. Sales reps use data and recommendations to have better conversations with customers.
“We want to be a trusted advisor to our customers, and this allows us to fulfill that role in a more consistent way,” Grande said.
Kodat can fill a certain void, Raut said. For example, some ERPs do not capture payment methods such as check, ACH, or card. Codat’s software will look at your bill and try to find that information, but it will tell you where it is making guesses or inferences.
“We use that information to say, ‘Hey, BMO, here’s a gap,'” Laut said. “This is how we think about payments, and that’s where we think AI is going to go. Showing us where the gaps are, clarifying what we inferred and how we got there. This allows the BMO team to have more advisory conversations with clients: ‘Hey, I guess you’re paying some of these vendors by check. How are you engaging with them today?’
About 15-25% of business payments are still made by check, Raut said.
“There are obviously a lot of issues with the checks,” he said. “One of the key issues is the speed of payments and control of working capital, which is a big focus for mid-market businesses. Cards give you more control over the timing of your payments.”
Raoult noted that check fraud is also common, and could get worse as fraudsters use AI images to create fake checks.
“This is a huge opportunity for businesses and banks to take their check payments and move them to more modern payments, whether it’s something even more outdated like ACH, or a card or part of a more modern payment rail.”
As BMO looks to expand its share with customers, playing an advisory role is important, Grande said.
“Now that we have this information powered and enriched by AI intelligence, we can go back to the customer and say, “Hey, you’re spending X amount of dollars on your business card program. We’re seeing that you’re likely spending a lot more than this.” Then we can incrementally increase benefits like cash back and reward points to help customers get the most out of it, she said.
Codat information also helps sales reps have money flow conversations with customers.
“This is where Treasury payments consultants add value to their clients. By moving this payment from cash or check, they can save or build on their working capital and benefit from things like delayed payments,” Grande said. “What we want to be able to do on a broader scale and in a highly customized and repeatable way is to enable financial consultants to provide their clients with insight into how they pay today and how changing it can improve their cash flow.”
For example, a customer may pay a supplier in full within 10 or 30 days of receiving an invoice.
“So we need cash flow to make that happen,” Grande said. “By making that payment a floating card, the supplier can get paid right away, but the customer himself has until the end of the billing cycle to make that payment. This allows the customer to spread out the flow of funds so they can manage other, more strategic business payments they need to make.”
Early returns
BMO is deploying Codat’s technology in the U.S. and plans to begin piloting it in Canada in the coming months, with a rollout by the end of 2026.
“So we’re fully on board,” Grande said. “We have customers using it to connect, and we are already starting to see some success stories from customers as well.”
Since BMO began using Codat’s software, several customers have shifted the majority of their payments from checks to cards.
“So we’ve increased our spending with BMO by about 45%, which is good for us, but it’s also good for our customers, because not only do they get the benefits that the BMO program provides, but they don’t have to stretch as much from a cash flow standpoint, and they save that much,” Grande said.
Reimer expects more banks to adopt this approach.
“The broader opportunity is not just AI-generated recommendations, but relationship-driven intelligence using permissioned data to understand where customers may have friction, whether it’s working capital, payment risk, uncovering fraud, or larger operational inefficiencies,” he said. “This improves customer relationships and increases revenue for banks by making customer conversations more relevant and contextual. Instead of generic B2B campaigns or regular banker check-ins, banks can tell customers what they see in the payment flow and suggest ways they can help. Additionally, by leveraging APIs to ingest this information, they can provide timely check-ins and AI By incorporating prompts, you can highlight your relationship with your bank.”
