Streamlining and evolving fraud investigations with AI

Cogniware, a global investigation platform provider to law enforcement agencies, needed to upgrade their existing system to better serve their clients in addressing these issues. Their existing system required manual effort for tagging and analysis, which wasn’t scalable for large volumes of data. This manual process was also prone to human error, potentially impacting accuracy.

Additionally, many of their clients’ investigators and analysts lacked the technical skills needed to fully utilize the platform’s advanced features. As a result, these employees spent time organizing data instead of focusing on analysis and decision-making, slowing down investigations and possibility raising threats to public safety.

Cogniware aimed to transform complex data analysis for law enforcement by integrating its Argos platform with IBM® watsonx.ai® and IBM watsonx.data® solutions. This partnership addresses the challenges of unstructured data and user accessibility, enabling fast and accurate investigations—all through advanced AI and data storage capabilities.

The watsonx.data solution helped store and manage vast amounts of data, particularly unstructured data. Meanwhile, IBM used watsonx.ai to enhance the platform with tools, such as natural language processing (NLP), machine learning (ML) and generative AI (gen AI). These additions enabled Argos to perform real-time data analysis, automatically find patterns and respond to queries asked in everyday language.

The implementation was a collaborative effort between Cogniware and the IBM Client Engineering and Technical Sales teams. These teams worked together to prepare data, train AI models and conduct user testing. To help ensure data security and prevent hallucinations, the teams implemented strong security protocols, trained AI models on specific data for contextual grounding and added fact-checking mechanisms. A key lesson Cogniware learned was the importance of continuous user feedback to improve gen AI features.

The integration of gen AI has fundamentally changed how investigators and analysts interact with the Argos platform. They can now “talk” to the system more naturally, eliminating the need for complex commands or premade reports. This ability frees investigative staff to spend less time organizing data and more time on analysis, which results in more efficient and comprehensive crime identification.

The integration has yielded significant, measurable improvements for Cogniware clients.

• Faster investigations: In a complex fraud case, what would have previously taken months to solve was resolved in days. The system’s ability to find subtle patterns and connections quickly led to a breakthrough.
• Improved accessibility: The new capabilities make it easier for nontechnical users to conduct complex data searches without needing a data analyst.
• Enhanced fraud detection: In the banking sector, the platform can now automatically find unusual transaction patterns that suggest financial crime, leading to quicker action and fewer financial losses.
• Positive client feedback: Clients have praised the ease of using natural language to ask questions and the significant reduction in time spent on data analysis. They’ve described “aha” moments where previously hidden insights were easily revealed.

Looking ahead, Cogniware plans to further refine their gen AI models for greater accuracy and context. The company also intends to expand the use of gen AI across other Argos modules, such as automated report generation and proactive threat intelligence systems. The partnership with IBM is part of a long-term strategy, with watsonx.ai and watsonx.data serving as the core for Argos’s analytics toolset to continuously improve data processing and user experience.

Cogniware (link resides outside of ibm.com) is a Czech-based software company and IBM Business Partner whose mission is to deliver innovative AI-based analytical products and elite professional services to clients worldwide. The firm’s core client base spans the public and private sectors from law enforcement agencies and national intelligence agencies to commercial banks and others.