Finastra has launched Repair Recommendations, a new AI powered capability within its AI OperatorAssist, aimed at helping banks fix payment exceptions faster as volumes grow. The company announced the feature at Sibos 2026 in Miami.
The cloud native feature identifies payment discrepancies, analyses the underlying causes and offers recommended corrections that are checked against the rules of payment schemes and networks, including Swift, Fedwire, SEPA, UPI and Nexus. It is meant for the teams that handle payment exceptions day to day, including wire room operators, exception and repair specialists and technical support staff, and Finastra says it will cut the time and effort they spend resolving problems.
Finastra said the feature will complement its Global PAYplus and Payments To Go products and described it as support for human decision making, with every recommendation subject to review and approval. By simplifying complicated repair scenarios, the company said, banks can handle exceptions more consistently, lower error rates, depend less on specialist expertise and bring new staff up to speed sooner. It pointed to rising demand for real time cross border capabilities, infrastructure modernisation and digital commerce as reasons banks need to manage higher payment volumes.
Barry Rodrigues, EVP of Payments at Finastra, said the feature is meant to help banks upgrade their payment operations by cutting resolution times and errors, and that AI can deliver immediate value as volumes rise by speeding up repairs, easing onboarding and strengthening resilience.
Repair Recommendations: What It Means
“With Repair Recommendations, we’re empowering banks to upgrade their payment operations, reducing resolution times and minimizing errors. As payment volumes increase, AI can help deliver immediate value by accelerating repairs, streamlining onboarding, and strengthening resilience. We’re helping banks to scale their operations, and deliver faster, more reliable services to their customers,” Rodrigues said.
Robin LoGiudice, Strategic Advisor at Datos Insights, said manual work remains a major operational risk for banks, with payment teams still spending long hours sourcing data and diagnosing exceptions.
“Manual processes continue to pose a major operational risk for financial institutions. Despite automation advances, payment teams still spend extensive time sourcing data and diagnosing exceptions, making operations costly and slow. AI-powered, validated, scheme-aware recommendations accelerate the gathering, extraction, and analysis of payment information, enabling faster exception resolution while improving efficiency and precision,” LoGiudice said.
For payment operations teams, a tool that suggests rule checked fixes could cut the time spent diagnosing each exception, and could help newer staff handle cases that once needed a specialist. For banks, it offers a way to take on more payment volume without adding the same amount of manual effort, though Finastra has not named any customers, shared performance figures or said when the feature will be available.
Digital Trade Outlook: What stands out is that Finastra is putting AI into the unglamorous part of payments, the repair queue, and not into the payment itself. Payment errors and disruptions are said to cost the industry billions each year, and a tool that knows the rules of Swift, Fedwire, SEPA, UPI and Nexus could be useful to banks that run several of them at once. The decision to keep every recommendation subject to human approval also matters, because it makes the feature easier for banks to adopt under their own controls. With no customers or results disclosed yet, the thing to watch is whether banks report measurable falls in repair times once it is in use.
Related on Digital Trade Outlook