
Automating Wholesale Order Processing
Automating Wholesale Order Processing
Prada Case Study
99.9%
99.9%
Reduction of Manual Formatting Time
Reduction of Manual Formatting Time
6
6
Distribution center files generated in seconds
Distribution center files generated in seconds




Project Overview
Industry
Fashion & Luxury Retail
AI Solution
Intelligent Excel Parsing & Data Transformation
Quick Brief
Prada's wholesale team spent 6 hours manually reformatting retailer order files for Oracle database upload. We built an AI system that automates the entire process in under 10 seconds.
Customer Pain Points
6 hours of manual work per order batch
Uploads rejected if even one cell is incorrect
Bottleneck in wholesale operations
What We Delivered
A browser-based AI application that automates the entire transformation process: Smart Parsing Engine Reads retailer Excel files, cross-references Prada’s internal style code database, and uses intelligent prefix matching to identify the correct variant codes. Automatic Transformation Handles every formatting requirement, from converting punctuation for Oracle compatibility to padding material and variant codes to the correct length. The system also organizes data by distribution center and sorts it by door, product, and size. Error Detection Flags mismatched store numbers and highlights the exact rows that require correction, allowing teams to fix issues in seconds instead of combing through spreadsheets manually. Output Generation Produces six distribution center files in seconds, delivering roughly 1,400 rows of clean, consistently formatted data. Every file meets Oracle’s upload requirements out of the box, eliminating manual prep.
End Results
No more spending entire afternoons on spreadsheet manipulation
Creates 6 distribution center files in seconds
6 hours → Under 1 minute
Introduction
Prada's wholesale team was spending hours on spreadsheets. Every time a major retailer like Nordstrom, Saks, or Harrods sent in an order, the team faced the same grueling process: manually extracting data from massive Excel files, reformatting it line by line, and preparing it for upload to their legacy Oracle database.
One wrong cell? The entire upload would fail. Six hours of work down the drain.
It was a bottleneck that slowed down their entire wholesale operation. With orders coming in from multiple retailers across different seasons, the team was spending 30+ hours annually on a task that should take minutes.
The Solution
We built an AI parsing system that transforms the entire workflow. The system takes three inputs:
1.The retailer's order file (eg. Nordstrom)
2.Prada's style codes reference
3.Door mapping (which products go to which distribution centers)
Then, in under 10 seconds, it generates six perfectly formatted Excel files (one for each distribution center) ready to upload directly into Oracle.
The system doesn't just parse data. It's smart enough to cross-reference style codes, handle non-standard formats, match variant codes, and even flag errors from the retailer's side before upload. If something's off, it tells the team exactly which row needs attention.
The team can now upload their files and in a blink of an eye have everything ready.

The Result
6 hours → Under 1 minute.
The wholesale team can now process orders the same day they arrive instead of blocking out entire afternoons. Less manual work means fewer errors, faster turnaround times, and happier retail partners.
Project Overview
Industry
Fashion & Luxury Retail
AI Solution
Intelligent Excel Parsing & Data Transformation
Quick Brief
Prada's wholesale team spent 6 hours manually reformatting retailer order files for Oracle database upload. We built an AI system that automates the entire process in under 10 seconds.
Customer Pain Points
6 hours of manual work per order batch
Uploads rejected if even one cell is incorrect
Bottleneck in wholesale operations
What We Delivered
A browser-based AI application that automates the entire transformation process: Smart Parsing Engine Reads retailer Excel files, cross-references Prada’s internal style code database, and uses intelligent prefix matching to identify the correct variant codes. Automatic Transformation Handles every formatting requirement, from converting punctuation for Oracle compatibility to padding material and variant codes to the correct length. The system also organizes data by distribution center and sorts it by door, product, and size. Error Detection Flags mismatched store numbers and highlights the exact rows that require correction, allowing teams to fix issues in seconds instead of combing through spreadsheets manually. Output Generation Produces six distribution center files in seconds, delivering roughly 1,400 rows of clean, consistently formatted data. Every file meets Oracle’s upload requirements out of the box, eliminating manual prep.
End Results
No more spending entire afternoons on spreadsheet manipulation
Creates 6 distribution center files in seconds
6 hours → Under 1 minute
Introduction
Prada's wholesale team was spending hours on spreadsheets. Every time a major retailer like Nordstrom, Saks, or Harrods sent in an order, the team faced the same grueling process: manually extracting data from massive Excel files, reformatting it line by line, and preparing it for upload to their legacy Oracle database.
One wrong cell? The entire upload would fail. Six hours of work down the drain.
It was a bottleneck that slowed down their entire wholesale operation. With orders coming in from multiple retailers across different seasons, the team was spending 30+ hours annually on a task that should take minutes.
The Solution
We built an AI parsing system that transforms the entire workflow. The system takes three inputs:
1.The retailer's order file (eg. Nordstrom)
2.Prada's style codes reference
3.Door mapping (which products go to which distribution centers)
Then, in under 10 seconds, it generates six perfectly formatted Excel files (one for each distribution center) ready to upload directly into Oracle.
The system doesn't just parse data. It's smart enough to cross-reference style codes, handle non-standard formats, match variant codes, and even flag errors from the retailer's side before upload. If something's off, it tells the team exactly which row needs attention.
The team can now upload their files and in a blink of an eye have everything ready.

The Result
6 hours → Under 1 minute.
The wholesale team can now process orders the same day they arrive instead of blocking out entire afternoons. Less manual work means fewer errors, faster turnaround times, and happier retail partners.
More case studies
More case studies
Deloitte
Predictive Market Intelligence
Predictive Market Intelligence
Industry
M&A Consulting
AI Solution
Co-Occurrence LLM
Quick Brief
Deloitte’s Emerging Technology and Global Markets team was looking to detect M&A activity ahead of the market. They were on the look out for ways to detect and gain access to the latest corporate M&A in real time.

Du Telecoms
Customer Service Analytics
Customer Service Analytics
Industry
Telecoms
AI Solution
Customer Service Analytics
Quick Brief
Du were looking to acheive full visibility on the operations of their call centres and retail stores by analysing the ongoing performance of customer service teams in both Arabic and English.

Bank of Poland
Document Translation
Document Translation
Industry
Banking
AI Solution
Multi-Language Document Translation
Quick Brief
NBP wanted a secure way of translating customer documents in 8 langauges through a centralised employee platform.

KukuFM
AI Audiobook Generation
AI Audiobook Generation
Industry
Media
AI Solution
Generative Text-to-Speech
Quick Brief
KuKuFM wanted to reduce the cost of using human voice artists and increase the speed of bringing new audio content to market.
© 2025 Perlon Labs LTD. All rights reserved.
© 2025 Perlon Labs LTD. All rights reserved.
© 2025 Perlon Labs LTD. All rights reserved.
© 2025 Perlon Labs LTD. All rights reserved.
© 2025 Perlon Labs LTD. All rights reserved.