Heritage · AI before AI
In 2019, well before the boom of generative artificial intelligence, we were applying neural networks to fast-fashion trend prediction. It is no longer our business: it is the proof, with a printed date, that we were doing AI before it was called that.
How it started
It all began when a client told us: why don't you use your financial trading systems to predict fashion trends?! Said and done. The link is direct: financial trends and fashion trends both depend on imitation among agents, and that is exactly what our models were built to detect.
The system recognised garments and colours from images with neural networks, clustered the colours, and produced trend indications for fast fashion — machine learning applied and in production, in 2019.
The story is told in a January 2020 article in the Italian trade press: the date that documents our "AI before AI". When artificial intelligence became the theme of the decade, we already had a real case behind us — and now that our clients are served by AI agents, that date is the simplest way to say we are not following a fashion. Not even in fashion did we follow it: we predicted it.
The evidence
FAQ
No: Redexe Fashion is no longer an active business. The page remains as historical documentation — with the January 2020 press article as dated evidence.
Because machine learning was in Redexe's production systems before the generative-AI boom: neural networks for garment and colour recognition and for trend prediction, documented by the press in January 2020. Now that every Redexe client is served by an AI agent, that date tells a continuity, not a conversion.
The method: agent-imitation models applied outside finance, and the familiarity with machine learning in production — the same you find today in Redexe's platform and AI agents.
Contact
If you are interested in how we apply AI today — dedicated agents, structured research, quantitative analysis — message Riccardo Donati on LinkedIn.
Redexe works strictly B2B.