A routing engine that removed a fifth of the fleet
Vehicle routing for Brazil's largest sugar producer — from whiteboard to production-ready in 16 months, across 40,000+ locations.
Brazil's largest sugar producer moved cane and equipment across a network of 40,000+ locations. Fleet was the dominant cost, and it was sized by experience, not by math.
A fleet sized by hand
Routing tens of thousands of locations by human judgment leaves vehicles idle and miles wasted every day — and no one could say by how much, or how to do better without more trucks.
Operations research, in production
Renan designed and led a Python optimization engine for fleet routing, started at Nitryx and carried through the Progress Rail (Caterpillar) acquisition. To make it cheap to run at scale, the team built in-house distance-matrix APIs instead of paying per external call, and integrated six real-time data sources.
He took it from concept to production-ready in 16 months, leading a team of six.
Fewer vehicles, lower cost, proven
The engine cut fleet needs by roughly 20% and saved about $60K per month in cloud cost through the in-house distance-matrix APIs — a routing capability the operation now runs on.
What BIS would do
If routing, dispatch, or fleet planning is a cost center for you, BIS builds the same kind of engine against your network and constraints — unbounded and exact where it pays — with the savings measured, not asserted. You can watch the method live on the transit demo.
Have a problem worth solving?
We start with a conversation, prove the value on a focused POC with your real data, then ship it — measurable ROI, not a model demo.