Skip to main content
Apolla AI
A beverage company

Consumer goods

Daily store visits per rep up 30%, distributor order errors down 80%

Beverage company: a distributor sales assistant that gets each rep into two more stores a day

Challenge

More than two thousand sales reps needed pricing policies, promotion rules, distributor inventory and order history every day, spread across five systems and countless group chats. Orders were keyed in by hand with a high error rate, and a new rep took three months to get up to speed.

The company was pushing deeper channel coverage. It wanted more visits and more output per store at the same time, without expanding the sales force.

Approach

We embedded a sales assistant in the reps' mobile app, turning lookups, calculations, ordering and reporting into a conversation.

  • Policy Q&A: pricing, promotion and rebate rules returned instantly for the distributor's tier and region
  • Smart ordering: orders captured by voice or photo, checked automatically for minimum quantities, bundles and stock, and turned into a draft order
  • Visit planning: a daily route and the products to push in each store, based on sales and inventory movements
  • Visit reports are generated automatically, so regional managers see structured data instead of chat screenshots

Architecture highlights

Integrated with DMS, CRM and the promotion management system through APIs, with policy data cached by region for weak networks. Voice orders use on-device speech recognition, and the model then handles intent parsing and validation.

Customer feedback

Reps no longer have to memorize policy, and they could never out-memorize the system anyway. Their time has gone back into the stores.
Vice President of SalesA beverage company

Want more detail on this project?

In a workshop we can share the architecture decisions, the pitfalls and the actual timeline.