Antoine Marchi

Antoine designs and maintains the part of Carouscale nobody sees: the pipeline that collects and classifies TikTok Shop carousels across six markets (the 5,000-view bar, the purchase-intent score, the market each post belongs to), and the AI that recreates those formats with each user's product, text baked into the image in the user's language. The collection rules are published on the methodology page, and the aggregates come out on the trends barometer.

Alongside Carouscale, he founded Cuvra, an agency that builds custom automations for small and mid-sized businesses, and he is a student at Epitech Nice. Before co-founding Carouscale with Sacha Jibril, he had already built an automated video generation system for a TikTok Shop creator doing €10,000 a month in revenue. That project is what led to the current tool.

His technical track record is detailed on his LinkedIn profile: security automation work at Docaposte, applied AI projects at Epitech, and a prize at the ETHGlobal international hackathon (top 20 and a Ledger bounty with the Iris Wallet project).

What Antoine builds on Carouscale

• The data methodologyThe 5,000-view bar, the intent score, market detection: the pipeline he designs.
• The trends barometerPublic aggregates per market, produced by the collection he runs.
• The featuresReplicate, the Create studio, alerts: what the in-house AI actually does.
• PricingPublic prices in euros, the counterpart of the infrastructure work.

The guides on this site are signed by Sacha Jibril, whose author page covers the selling side. Antoine answers for the numbers: every statistic shown on Carouscale comes out of the pipeline described on the methodology page, known limits included.