Post-trade is the place for agentic AI workflows

In the next few years, most post-trade operations analysts won’t be bogged down with processing exceptions. Instead, they will supervise agents, configure workflows, manage counterparty relationships at a strategic level and handle cases that require human authority, according to Yogesh Shenai, director of product management at Smartstream. Supporting the smooth running of the ongoing global expansion of T+1 settlement in the post-trade space is precisely the type of ‘problem’ agentic AI has been waiting to solve, he says.

Exception queues weren’t prioritised by settlement urgency, so the most time-critical breaks weren’t always the ones getting attention first, Shenai goes on to say. However, when settlement failed, the downstream workflow of counterparty notification, partial settlement decisions and penalty tracking remained largely manual, he says. Meanwhile, standard settlement instruction (SSI) applications unveiled gaps in validation and confirmation timing under the compressed window, he adds.

According to Shenai, automating routine exception handling reduces manual touch rates by 60-80% across in-scope workflows. Meanwhile, an investigation that takes a skilled analyst 30 to 45 minutes can be completed by an agent in less than two minutes, he adds.