Client:
Suzano International
Suzano, established in 1924 by Leon Feffer, is a leading Brazilian pulp and paper company headquartered in Salvador, Brazil. Under the leadership of Chairman David Feffer and CEO Walter Schalka, Suzano has grown to become the largest paper and pulp company in Latin America and a global leader in eucalyptus pulp production, serving over 80 countries. Operating within the pulp and paper industry, Suzano employs approximately 37,000 individuals. In 2023, the company reported revenues of US$8.5 billion and a net income of US$ 869.9 million.
Suzano's product portfolio includes a variety of raw materials such as pulp, fluff pulp, micro-fibrillated cellulose, and lignin. Additionally, they offer paper and packaging solutions alongside hygiene products. The company maintains a strong commitment to sustainability, focusing on innovative and eco-friendly solutions to meet global demand. Suzano's extensive operations span all regions of Brazil and several international locations, including Argentina, Austria, Canada, China, Finland, Israel, and the United States. Nanonets is primarily involved with Suzano in Austria.
- Location: Vienna, Austria
- Year of Establishment: 1924
- Leadership:
- Leon Feffer, Founder
- David Feffer, Chairman
- Walter Schalka, CEO
- Industry: Pulp Production and Paper Manufacturing
- Employee Count: 10,000+
- Revenue: US$ 1-10 Bn.
The Challenge
All Purchase Orders placed with Suzano in Europe are processed by the Sales Teams. The relevant data points are entered into SAP which is the central ERP software that they use. Once done, a Sales Order is created to be processed downstream with supply and delivery confirmation.
As stated, due to varying formats and document types, this process was manual and needed optimization. But before looking at the customised solution Nanonets offered the team at Suzano Europe, let’s take a deeper look at the process in order to understand it better:
- End-customers create a Purchase Order which is shared with the sales teams at Suzano. The challenge is accurately identifying and extracting relevant data points from the hundreds of templates for the purchase orders across 70+ end-customers, not to mention the different document types, such as, PDFs, Emails, Excel Sheets, Printed scans of Excel Sheets, etc.
- Once a Purchase Order is received, the relevant data points are manually entered into SAP. After this, a Sales Order gets generated which is utilised downstream for record-keeping as well as delivery confirmations.
The most time-consuming step was gathering and homogenising the data for facilitating data entry into SAP. Once optimised, the idea was to reduce the time taken to enter data manually significantly reducing the turnaround time.
OCR Technology has obviously been available for 10-15 years. We had been testing different solutions but the unique aspect of Nanonets, I would say, was its ability to handle different templates as well as different formats of the document which is quite unique from its competitors that create OCR models based specific to a single format in one automation. So, in our case as you can imagine we would have had to create more than 200 different automations.
The Solution
Nanonets understood the entire manual process and rose to the challenge of automating it. We ingest Purchase Orders directly at the source, which is Gmail or One Drive, automatically extract the relevant data points, format them, and export them as Excel Sheets. Then, the team leverages VBAs and Macros to automate data entry into SAP.
Let’s take a look at the automated process with Nanonets:
- Import: Nanonets ingests the incoming Purchase Orders directly at the source, which is Gmail or One Drive. For all other document types, such as PDFs, Excel Spreadsheets, hard copies of Excel Spreadsheets or even screenshots, there is an option for manual upload.
- Data Extraction: The AI-powered custom OCR model at Nanonets extracts the relevant data points from the Purchase Orders, including relevant customer details, such as, name, shipping and billing address, along with a list of requested items, including, name, description, quantity, etc. in a tabular form.
- Post-Processing, Mapping and Lookups: This data then flows through a number of post-processing formatting steps such as finding and replacing characters, removing multiple instances of currency symbols, etc. to clean up the extracted data. We also create additional data points that work as “lookup codes” based on shipping and billing addresses for efficient vendor mapping within SAP.
- Approvals: There is an additional approval step in place that ensures all data points are extracted correctly with no oversight.
- Export: The clean, formatted data is then extracted as Excel Spreadsheets. The team at Suzano utilises VBAs and Macros to automate data entry into SAP.This is a solution which requires minimal human intervention from the team for data entry. We will also be automating data export to SAP in the coming months.
This is a solution which requires minimal human intervention from the team for data entry. We will also be automating data export to SAP in the coming months.
Another unique aspect was that it is accessible. It allows other businesses to test it out and operate themselves, as opposed to other competitors which require advanced programming skills and some don’t even allow you to test it yourself, which in our case is much appreciated because it allows contribution from the team and operational flexibility as well as visibility into who is doing what and when? The solution delivered exactly what was advertised, and more, unlike other competitors.