AI Finds and Analyzes Images and Videos to Identify Financial Risk

Summary:

A company leveraged AI to analyze images and videos for detecting potential risks. By automating the extraction of critical data from visual and textual content, the company improved its efficiency in risk assessment and compliance monitoring for its clients.

Client: 

KPMG is a global network of professional services firms providing audit, tax, and advisory services. With a presence in over 140 countries and territories, KPMG serves businesses, governments, and non-profit organizations. 

Problem Statement: 

KPMG needed a robust solution to analyze large volumes of visual and textual data from images and videos to identify compliance issues, fraudulent activities, and other risks. Manual analysis was time-consuming, error-prone, and insufficient for handling the scale and complexity of modern data sources.

Results: 

• Reduced time required for risk analysis within two days.
• Achieved higher accuracy in identifying potential risks from images and videos. 
• Improved compliance.
• Streamlined compliance monitoring and reporting processes, saving significant operational costs.

AI Solution Overview:

KPMG integrated Azure Cognitive Services to enhance its AI capabilities across various functions. By incorporating Azure Cognitive Service for Speech into its Magna platform, the company automated the transcription of conversations, making them searchable through detailed transcripts. Azure Cognitive Service for Language helps KPMG extract and organize text efficiently from documents.

Using Azure Computer Vision, KPMG improves its ability to detect and analyze visual content, including leveraging optical character recognition (OCR) APIs to extract text from images and videos. Serving banking clients globally, KPMG deals with a workforce and clientele that often communicate in multiple languages, including regional slang and informal expressions. To address this, the company implemented Azure Translator within Magna, enabling both prebuilt and customized translation solutions.

To consolidate its capabilities further, KPMG adopted Azure Cognitive Search, which provides contextualized searching of unstructured data within customer ecosystems and enhances the accuracy of search results. Additionally, KPMG utilizes semantic search in preview mode—advanced machine learning models trained on extensive Bing query data—to deliver highly nuanced and contextually relevant search results across its indexed data.

References: 

1. KPMG helps banking customers identify financial risks by unifying data with Azure AI. https://www.microsoft.com/en/customers/story/1554364202574065104-kpmg-professional-services-azure-ai

Industry: Financial Services and Emotion detection

Vendor: Microsoft

Client: KPMG

Publication Date: 2022