Back to Insights
Technology
Feb 20, 2025 · 5min read
Broker estimates. Visual assessments. Negotiated grades. The old grading model systematically undervalues material and exposes buyers to quality risk. ML changes the equation.
The secondary materials economy operates at enormous scale — but with the infrastructure of an earlier era. Most market participants accept this as the cost of doing business. Veloce was built on the belief that the cost is structural and solvable.
The fragmentation isn't random. It emerges from the fact that platforms solve for their silo — auctions, or compliance tracking, or raw material procurement — and leave the connections between them to brokers, phone calls and manual coordination. Each silo appears to work. But the cost of the connections between silos is where the value leaks.
Veloce's design principle is to own the full stack: ML grading that creates objective data before discovery, a field network that creates physical infrastructure before digital, and three verticals that share the same material flow, compliance record and buyer network. The result isn't three products on one login. It's one platform where the output of one vertical feeds the next.
