An international payments giant, mid-merger · Payments technology · AI & agentic commerce
A defensive war of merger integration at home, an offensive war against AI-native challengers ahead — and the discovery that the same build wins both.
Client identities withheld by design. Engagements delivered as lead across previous consulting roles; figures are each engagement's findings and committed design targets.
The situation
Mid-merger, the processor had promised the market a multibillion-dollar synergy program — a defensive war of platform consolidation and cost-cutting. Meanwhile AI-native challengers were publicly building toward agentic commerce, where software negotiates and pays on a customer's behalf. On the ground, the patchwork showed: a merchant's point-of-sale data didn't match her settlement ledger, disputes meant hunting receipts across portals, and internal ops staff spent most of their time as human glue between systems that didn't reconcile. A rival processor had recently run the same merger playbook fix-only — and watched tens of billions in market value evaporate.
What the diagnosis found
Both sides of the counter, same wound
The merchant's friction and the ops team's friction were mirror images: she reconciled her sales against opaque statements by hand; he reconciled the same mismatched ledgers from the inside — 60% of his time as human glue. One data seam, two burnouts.
Friction priced in churn
Every unexplained fee and week-long dispute pushed merchants toward unified-platform challengers. The patchwork wasn't just an operations cost — it was actively feeding the competitor's pipeline.
The cautionary tale had a name
A peer processor's fix-only merger — cut costs, under-invest in experience — stalled its platform, bled merchants, and erased tens of billions in value. The market had already run the control experiment.
The fix and the lead are the same build
The unified data layer needed to hit internal savings targets is the identical foundation agentic commerce requires. Framing them as competing priorities was the strategic error; they are one engine with two applications.
The redesign
An AI reconciliation engine
One unified data layer matching every ledger across the merged estate — 99.9% of transactions matched automatically, with only true anomalies escalated. The month-end week becomes minutes; the human glue becomes analysis.
The same dispute tool, inside and out
A GenAI dispute co-pilot that auto-gathers order, signature, and settlement evidence in seconds — given first to internal ops, then to the merchant herself. The internal fix becomes the external sticky feature.
An agent layer on top
A growth twin that benchmarks each merchant and proposes same-day promotions, a silent resolver that pre-empts failures before morning, and a compliance shield that blocks risky transactions in real time — not weeks later.
Fix-to-lead, by design
Every engine built for the synergy program is designed with a customer-facing twin, so the defensive war funds the offensive one — and the platform graduates from processing payments to anticipating commerce.
What the blueprint committed to
99.9%
auto-matched transactions — anomalies alone reach a human
1wk→10m
month-end reconciliation, from a week of spreadsheets to minutes of review
2-in-1
every fix engine doubles as a lead engine for agentic commerce
The systems you build to fix yesterday's mess are the same systems that win tomorrow's market — if you design them facing the customer, not just the cost line.
The first principles at work
Recognize your own journey in this?
Tell Exie what feels similar — the same diagnostic method maps your version of it. Free, and you keep the map.
Next case · A wealth-management firmDormant isn't gone.