The challenge
Ensure effectiveness and reliability in automating a process that eliminates manual intervention in the daily management of filed complaints, within a workflow centered on reviewing documents individually, where data from an email message and its attachments is extracted, indexed, and classified by reason, channel, and product.
The solution
Google's AI Platform enabled our developers, data scientists, and data engineers to implement a solution with extensive GPU processing power, based on an ML model for classifying complaints through instances running Python and Jupyter Notebooks and using Vision API OCR for keyword searches within documents.
Project results
An intelligent, reliable processing solution for classifying filed complaints: serverless, cost-effective, and highly scalable. It ensures data extraction from large volumes of unstructured documents, eliminates manual classification, maintains data privacy, boosts productivity, and reduces response times.
SFC found a system with advanced analytics capabilities for effective management, beyond what is currently available in on-premises solutions, providing the reliability, efficiency, security, and scalability needed to meet mission-critical business requirements at a lower cost.
About the Colombian Financial Superintendence
The Financial Superintendence of Colombia aims to supervise the Colombian financial system to preserve its stability, security, and trust, as well as to promote, organize, and develop Colombia's securities market and protect investors, savers, and policyholders.
Industry: Financial services. Location: Colombia.
