Rocoto automates the conversion of incoming orders from WhatsApp, voice notes, photos, and email into structured delivery notes (albaranes) using each business's own product catalog, pricing, and customer history. The platform applies six validation checks per order, resolves colloquial product names to catalog items, and exports the final data to Excel, Sage, or via API. It targets food-service suppliers who currently lose around 10 hours per week manually transcribing orders.
Funding
Funding not disclosed
Founders
Product
Problem
Food-service suppliers and distributors receive a high volume of orders through informal channels like WhatsApp messages, voice notes, photos, and email, which must be manually transcribed into their order management systems. This manual process is time-consuming, error-prone, and scales poorly, consuming roughly 10 hours per week for a typical provider and delaying order processing.
Solution
Rocoto is an AI-powered order-processing platform that automatically reads incoming orders from WhatsApp, voice notes, photos, and email, then converts them into structured delivery notes (albaranes) using the business's own product catalog, pricing, and customer purchase history. The system interprets colloquial product names and variants—such as "la paleta" or "el york del bueno"—and maps them to the correct catalog items, including options like "entera," "sin cabeza," or "de pincho." Each order passes through six validation checks before being finalized, and any order with uncertainties is flagged and surfaced to the team with the specific reason for review. The final output is exported directly to Excel, Sage, or via API, ensuring the team works in the tools they already know.
Target Audience
Primary customers are food-service suppliers, wholesalers, and distributors in the hospitality sector that receive orders via WhatsApp, voice notes, photos, or email and need to streamline their order-to-delivery workflow.
Features
- Multi-channel order ingestion from WhatsApp, voice notes, photos, and email into a single unified inbox
- Natural language processing that resolves colloquial product names and abbreviations to the correct catalog items
- Support for product variants and modifiers (e.g., "entera," "sin cabeza," "de pincho") with automatic line-item selection
- Six automated validation checks per order to catch errors before finalization
- Automated flagging of uncertain orders with the specific reason, prioritized for human review
- Direct export to Excel, Sage, or via API for seamless integration with existing workflows