Vue.ai launched in 2016 as retail’s AI flagship: it tagged catalogues, staged on-model imagery, and built a virtual dressing room that put a coat on a body like the shopper’s own. The pages it builds now are about loan approvals, insurance claims, and figures pulled from scanned documents. That drift from the fitting room to the loan file is a verdict on whether retail-AI margins can hold a category leader.

The company’s own words track the retreat. Its parent, Mad Street Den, was backed by Avatar Growth Capital and Sequoia and, for a time, a name that stood in for retail AI itself. In January 2023 it took a $30 million Series C and promised to “double down” across finance, insurance and logistics while “holding firm as the Leader in Retail AI.” A firm that leads a market seldom needs to say so; and it never says so while listing the markets it would rather be in.

The logic sits in the margins, not the mission. Fashion hires the machine to flatter, to stage a dress well and lift add-to-cart by a fraction. Finance wants something duller, and pays more for it: read the loan file, run the checks, retire the clerk. A retailer cuts the first in a soft quarter and scarcely feels it. The second is a saving a bank can put a number to, and a number is the only case a chief financial officer reliably funds.

Look at the number the company chose to boast about. Its dressing room, one case study reported, produced a 1.5x lift in add-to-cart for the sessions that used it. That is a real result, and a small one, the sort a merchandiser files under “experiment.” A bank does not run experiments on whether to process its loans.

Retail AI carried a second flaw: it kept getting cheaper. Google launched its own virtual try-on inside Shopping, and Vue.ai published a response — titled “Utterly Clueless”, a Clueless film riff that called Google’s training dataset “unmatched canvas” before arguing Vue.ai’s product went further. You do not publish a differentiation pitch against a rival you are comfortably ahead of.

“Horizontal AI platform” is the phrase vendors reach for when one vertical stops paying. It promises breadth; it usually confesses that the first market did not hold. Vue.ai spoke the phrase fluently by the end. The confession was in the fluency.

In March 2025, the self-styled Leader in Retail AI sold to a digital banking infrastructure company in a distress sale first reported by Moneycontrol, cash and stock, for a fraction of the more than $50 million it had raised.

The buyer was M2P Fintech, which runs banking and payment rails. It did not want the dressing room. It wanted the machine that had learned to read loan files — the one thing Vue.ai had built that a bank would pay for. The dressing room came free with the deal.

Vue.ai did not fail at fashion. It led at fashion, then learned that leading a market this thin is a patient way to lose. The dressing-room mirror still hangs on the website; behind it, the firm is reading a stranger’s mortgage.