How Wonga Uses PyMuPDF to Power High-Performance Financial Document Processing

Company Overview

Wonga is a leading financial technology company specializing in short-term digital lending. Founded in 2011, the company offers fully automated, online consumer loans, leveraging cutting-edge technology to deliver fast, transparent, and flexible financial products. Wonga employs a microservices architecture and advanced data science workflows to power its operations, from automated risk assessment and affordability checks to document processing and customer onboarding.

The Situation

Wonga needed a high-performance solution for extracting structured information from financial documents to support its data science workflows. The company processes large volumes of PDFs—including bank statements, payslips, and income verification records—as part of its automated lending pipeline. These documents arrive in a wide variety of layouts, and extracting accurate, position-aware text from them is essential. The team required word-level extraction with precise positional coordinates to build reliable parsing logic on top of complex PDF structures. Speed was equally critical: any bottleneck in extraction would cascade into delays across the entire document processing system.

The Solution

Wonga selected PyMuPDF to power its automated document processing systems. PyMuPDF stood out for its:

• Speed—close to an order of magnitude faster than alternative libraries in benchmarking
• Accuracy in word-level text extraction with precise positional coordinates
• Ease of use and straightforward integration into existing workflows

The performance advantage was critical for Wonga, as the company processes large volumes of documents and requires both efficiency and precision. With PyMuPDF, the team built a robust document intelligence layer that extracts structured data from financial PDFs and feeds it directly into automated decision-making systems. The precise coordinate data provided by PyMuPDF enabled layout-aware parsers that correctly interpret documents regardless of formatting inconsistencies.

PyMuPDF has become a foundational component in Wonga’s document intelligence stack, enabling the company to build scalable, production-grade PDF processing systems.

The Results

By integrating PyMuPDF, Wonga achieved significant performance gains in its automated document processing pipeline. Extraction speeds improved by close to an order of magnitude compared to previous solutions, directly accelerating the company’s lending workflows. The accuracy of word-level extraction with positional data enabled the team to build reliable, production-grade parsing logic that handles the full range of financial document layouts encountered in real-world operations. PyMuPDF now underpins Wonga’s scalable document intelligence systems, supporting the company’s growth and its mission to expand financial inclusion through technology-driven lending.