Talk to us
All solutions
All industries
All teams
How we work
Assess, build and run, with one accountable team before and after go-live.How we work 2 weeksfrom scattered ideas to a ranked, delivery-ready plan.The AI Assessment Data scientists and engineers shipping to production, then maintaining it.Data and AI engineering Your data stays yours, governed from the first day.Trust and governance
Featured stories
See all success stories
4.2 hto collect 2,800 financial reports with fifty agents in parallel, down from 210 hoursRead the story 35%more coverage of points of sale for Heineken field teamsRead the story 50k+cable designs tested virtually for CoficabRead the story 800K+calls a month transcribed, summarised and classified at NOSRead the story
Resources
Information Hygiene: The Security Discipline That Matters More Than EverRead the blog Artificial Intelligence Made in Portugal: ‘We grew more than 60% in one of the most competitive markets in the world,’ says DareData Partner.All news
Company
2019Founded in Lisbon by engineers. Today we build and run AI for large companies across Europe.About DareData 130+engineers, data scientists and principals in Lisbon, Porto and LondonMeet the team We are hiring engineers who want to build AI that runs.See open roles 50+customers served. Tell us about yours.Talk to us

Sogrape turns distributor emails into SAP orders, matching customers with 94% accuracy

Distributor order emails turned into SAP orders

Industry
Consumer & Retail
Solution
Automation

Distributor order emails turned into SAP orders

Industry
Consumer & Retail
Solution
Automation

Sogrape turns distributor emails into SAPorders, matching customers with 94% accuracy

Client
Sogrape
Industry
Consumer & Retail
Solution
Automation
Technology
Gen AI, GenOS
Team
2 DareData engineers
94%customer correspondence accuracy
82%product match accuracy in initial tests, still improving
~0.5 FTEfreed by automation

Product matching from initial tests, with further improvements expected

Sogrape's distributors place orders by email, and customer care keyed each one into SAP by hand, using rules that lived in people's heads. GenOS Supervisor now extracts the order from the email and its attachments, matches it against SAP master data, and hands the operator a filled order to approve. Fifteen minutes of keying per order becomes a review.

With DareData, Sogrape:

  • Extracts order details and SKUs from email bodies and PDF attachments
  • Matches clients, products and pallets against SAP master data
  • Reaches 94% customer correspondence accuracy, and 82% product match accuracy in initial tests
  • Gives operators an editable order to approve before anything reaches SAP
  • Turns undocumented business rules into deterministic code
  • Covers distributors across Europe, the US, Australia and South Korea

The challenge

Every order keyed in by hand

Sogrape ships wine to distributors across Europe, the US, Australia and South Korea, at around 300 shipments a month. Orders arrive by email, in the body or as attachments, and each one has to become an order in SAP.

Fifteen minutes per order

Customer care keyed every order into SAP by hand. Each one took about fifteen minutes, with room for error at every step: the wrong client, the wrong product code, the wrong pallet.

Rules that lived in people's heads

The business rules for handling orders were never written down. They existed in the experience of a few operators, which made them hard to share and impossible to apply during a staff absence.

No room to grow

As the number of distributors grew, so did the manual work. The process had no way to scale without adding more people to type.

The solution

A filled order, ready to approve

The system runs on GenOS Supervisor, which orchestrates the flow from the moment a distributor’s email arrives to the moment an order is created in SAP.

Each incoming email is read in full, including inline notes and PDF attachments. Generative AI and Document AI extract the order details and SKUs, and turn free-form text into structured order data.

The hardest part was not the extraction but the rules. How a given distributor’s order should be handled lived in operators’ heads, so the logic was worked through with customer care and written as deterministic code. Those rules run before the language model, which keeps its decisions simpler and the extraction more accurate.

The structured order is then matched against SAP master data: the client, client-specific product identifiers, delivery addresses and pallets. SAP’s own field checks, such as shipment dates and delivery blocks, are validated before submission, and a single email can be split into several SAP orders when it needs to be.

Nothing reaches SAP without a person. The operator sees the order in an editable table, can edit, save or cancel it, and approves it before submission. The system is positioned as augmented work rather than replacement: the team reviews the cases that need judgement instead of typing every order. It runs on Google Cloud, on GKE.

How it works

From data to decisions, in 4 steps.

  1. 01Read the emailGenAI and Document AI extract SKUs and order details from the email body and attachments.
  2. 02Apply the rulesDeterministic business rules run before the language model to keep extraction accurate.
  3. 03Match against SAPClients, products and pallets are matched automatically against SAP master data.
  4. 04Approve and submitThe operator edits or approves the order in GenOS Supervisor, which then creates it in SAP.

The impact

From keying orders to reviewing them

  • 94%customer correspondence accuracy
  • 82%product match accuracy in initial tests, still improving
  • ~0.5 FTEfreed by automation

Order handling moves from fifteen minutes of keying to near-automated. Operators no longer type orders: they review a filled order and approve it, and only then does anything reach SAP. That is equivalent to about half a full-time role freed for other work.

Customer correspondence accuracy reaches 94%. Product matching reached 82% in initial tests, with further improvements expected as the matching rules mature.

The rules that used to live in operators' heads are now written as code, so the process no longer depends on who is in the office. The same pattern applies to any business that receives orders by email and keys them into an ERP.

The team

The people who built it.

  • Daniel ZuritaSenior Data Engineer & Project Manager
  • Luiz LeiteData Scientist
The solution behind this storyData and AI engineeringSee how we build it

Your AI story, next.

Tell us what you want AI to do for your business. We design, build and run it with your teams, and measure it like the stories above.

  • 01Use cases chosen for value, not hype
  • 02Systems built into your stack and data
  • 03Run, monitored and improved after go-live
Talk to usBook an assessment

More from production.

All success stories
Rui MendesHead of GrowthHave a similar problem? Book 30 minutes.No slides. We look at your use case together.Pick a time