Automated order-to-route assignment with 1C
Challenge
Every day operators went through the stream of orders by hand and decided which driver or which fixed route would deliver each point. Orders arrive all day, the plan has to be rebuilt constantly, and a bad assignment means extra mileage and missed deliveries.
Solution
We built an auto-assignment engine on NestJS on top of the 1C data. It learns from 56 days of delivery history to predict the driver for a point, and where history is thin it falls back to contractor rules (fixed driver, route, or day-of-week market run). Every 5 minutes until a 17:00 cutoff the plan is rebuilt and reconciled — new rounds created, changed ones patched, empty ones removed. Finished drafts are written back into 1C over OData, and a separate shadow mode compares the system's prediction against what operators actually did each day.
Result
Rounds are assembled automatically in real time, leaving operators to review and confirm the plan by 17:00 instead of building it from scratch. Assignment accuracy is measured in shadow mode and shown in a report — point-level learning combined with contractor rules keeps predictions stable.
Want to talk through your problem?
Free 30-minute audit — we'll show what can be improved and what it costs.
Request an audit