Machine Learning

Tradeshift’s Next-Level AI/ML Puts Payables Departments in Control

E-invoicing and accounts payable automation market leader Tradeshift released a series of innovations to Ada™, a next-level artificial intelligence and machine learning layer embedded on the Tradeshift platform. Accounts payable teams using the Tradeshift® Pay solution now have access to a powerful set of tools designed to help them visualize the impact and effectiveness of Tradeshift’s payables automation capabilities compared to human-operated processes.

The latest iterations include an Automation Dashboard™ feature that measures Ada’s performance against person-led processes and offers predictive modeling showing how the technology can maximize outcomes, like gained efficiency and cost-savings. Ada’s invoice coding functionality is constantly learning from valid historical transactions – even when it is not in use – to achieve complete, accurately coded documents that are ready for approval and posting.

Understanding that most are not comfortable with fully handing over the keys to Artificial Intelligence, customers can select one of three presets; conservative, balanced, progressive. The presets are optimized to scale AI according to the customer’s confidence in leveraging the technology. Early adopters of the system have reduced manual coding interventions by an average of 60 percent, with a top-performing customer saving nearly an entire week’s worth of work (35 hours) in a single month.

Tradeshift has also added a rule engine designed to complement Ada’s machine learning capabilities by providing users with a range of options to help them customize the system according to their unique specifications. Ada Rules provides users with an additional layer of flexibility and control over AI deployments by enabling users to apply special conditions when invoices meet a set of given criteria, such as atypical invoices from infrequent vendors, unusually large invoice values, or exceptions to international compliance requirements.

“We don’t believe in building ‘black box’ technology,” says Lloyd Humphreys, Tradeshift’s principal product manager for analytics and artificial intelligence. “Trusting machine learning with something as important as ensuring your company’s suppliers get paid is not something that should be done behind a curtain. That’s why we believe that the Automation Dashboard’s transparency and ability to ease into Ada is the key to supporting exception-free payables departments and teams that can be liberated to do more meaningful work.”

Customers of the Tradeshift Pay solution can process 10s of thousands of invoices per month, a labor intensive and unrewarding task for overburdened payables teams which are prone to mistakes and human error. According to Ardent Partners, payables teams spend an average of 24% of their time handling invoice exceptions, while a separate study concludes that 3.6% of invoices entered manually include an error caused by data entry.

“The best processes to automate are tedious, like coding invoices for a successful ERP ingestion. Even the most passionate data entry clerk would tell you there are more valuable things they could be doing with their time. Ada is the ultimate partner for payables departments looking to move up another level” says Raphael Bres, chief product officer at Tradeshift.

For more such updates and perspectives around Digital Innovation, IoT, Data Infrastructure, AI & Cybsercurity, go to

Related posts

Digma launches Continuous Feedback Platform for Code


ML-driven drug discovery company, insitro Raises $400M

Business Wire

Chris Ramsey appointed as CEO of Cybercheck

PR Newswire