Extracting Data at Enterprise Volume
Imagine being asked to build a database from more than 100,000 PDF invoices, each one formatted differently and containing dozens of line items, serial numbers and fiscal details.
Now imagine doing that without pulling teams out of daily operations and without compromising regulatory accuracy.
For OXXO, this was not a hypothetical exercise. It was the next logical step in managing fixed assets on a national scale.
“Manually, it simply wasn’t possible,” said Patricia Fabila, transformation, administration and finance manager at OXXO. “There was no practical way to extract that information without automation.”
When Scale Demands Structure
As part of FEMSA (Fomento Económico Mexicano), a multinational conglomerate and the world’s largest Coca-Cola bottler, OXXO operates more than 25,000 convenience stores across Mexico. Each store opening and equipment update generates fixed assets such as refrigeration units, shelving and point-of-sale equipment. Every asset is backed by an invoice that must be retained for accounting, regulatory and insurance purposes.
Over time, OXXO used Laserfiche to digitize and centrally manage all of its fixed-asset invoices, establishing a trusted system of record. Today, that system contains more than 1 million documents and continues to grow by approximately 1,500 invoices each month. While this foundation supported audits, insurance claims and regulatory requirements, OXXO saw an opportunity to unlock even more value from the information.
“We had all the invoices digitized in Laserfiche,” Fabila said. “What we needed next was the ability to work with the information inside those documents.”
That need became more pressing as OXXO worked to reconcile internal asset records with information reported to Mexico’s tax authority, SAT. While SAT maintains its own record of issued invoices, OXXO needed to confirm that those records aligned with its internal data across years of historical documentation. At that scale, searching PDFs one by one was no longer viable.
Turning Invoices into a Database with Smart Fields
OXXO partnered with Expert Data to extend its existing Laserfiche environment using Smart Fields, an AI-powered data extraction tool. The objective was to extract structured data from invoices at scale and convert historical documents into a dataset that could be queried, validated and analyzed.
“OXXO had a clear vision for how it wanted to use its data,” said Gasi Fayad, director at Laserfiche solution provider Expert Data. “Smart Fields made it possible to extract and structure information on a level that would not be feasible manually. The result is a database that continues to deliver value over time.”
Smart Fields uses AI to automatically capture and apply metadata from documents, even when layouts vary. This flexibility was critical, as OXXO receives invoices from a wide range of suppliers using different formats.
Rather than stopping at invoice-level metadata, OXXO designed the solution to capture detail at the line-item level. Many invoices contained dozens or even hundreds of individual assets, each of which needed to be represented as a distinct data record.
Using Smart Fields, OXXO extracted structured data from more than 100,000 historical invoices. Because each invoice could generate multiple line items, the initiative produced millions of individual data rows. Each one represents a specific asset with its own serial number, value and reference back to the original invoice. OXXO organized this data into a fully queryable database with defined rows and columns.
Smart Fields captures and structures data including:
- Invoice numbers and SAT fiscal UUIDs (universally unique identifiers)
- Supplier and fiscal information
- Invoice totals and accounting values
- Individual asset descriptions and serial numbers
- Line-item values tied back to each invoice
Once structured, OXXO could search, filter and compare this information across years of historical data.
“With Smart Fields, we moved from having invoices in a repository to having a real database,” Fabila said. “Now we can see exactly what was purchased, item by item, with data we can validate and analyze.”
Proactive Insight Across a Nationwide Operation
With Smart Fields in place, OXXO shifted from document searches to data queries. Teams can now locate assets by store, serial number or fiscal identifier across millions of structured data rows, rather than manually reviewing large sets of invoices. When regulatory authorities request information, OXXO can respond using a centralized dataset that reflects both historical and current operations.
The database also supports internal validation. Teams can compare accounting entries against invoice data, and verify asset records using detailed, line-item information.
While the project was not designed to generate direct cost savings, its value lies in preparedness and risk reduction. “Laserfiche helps us respond faster and with more confidence to regulatory requirements,” Fabila said. “That peace of mind is extremely valuable for an organization of our size.”
Expanding the Historical Record
OXXO, together with Expert Data, designed the Smart Fields initiative to grow over time. With a structured data model already in place, the team is now evaluating how far back to extend data extraction across its historical archive.
The organization can selectively process additional years of invoices as regulatory or operational needs evolve, continuing to strengthen the database with each expansion.
“Laserfiche plays a strategic role in how we manage information,” Fabila said. “It gives us confidence that we can support regulatory requirements, operate efficiently and make decisions based on reliable data.”
