New Way Now: ComplyAdvantage keeps financial crime at bay with Google Cloud скачать в хорошем качестве

New Way Now: ComplyAdvantage keeps financial crime at bay with Google Cloud

Featured in this video: Mark Watson, Chief Product Data and Technology Officer at ComplyAdvantage Executive summary: ComplyAdvantage, an AI-driven risk intelligence platform, partnered with Google Cloud on its mission to eliminate financial crime. Using Google Cloud’s planet-scale infrastructure and Security Command Center, ComplyAdvantage can support its multicloud strategy while meeting the financial sector’s demanding resilience requirements — achieving 4x more region coverage and maintaining over 99.99% availability for its customers. Google’s Gemini models are also enhancing its fraud and anti-money laundering risk detections: ComplyAdvantage built an intelligent tool chain around Gemini that can screen eight million articles a day and process 1.7 million adverse media reports across public data sources, delivering 2x more accurate classification and subclassification. Plus, alert volumes have dropped by up to 85% with Gemini’s agentic auto-remediation. Crucially, the team can now use natural language to adapt models for emerging risks in the same day — a massive leap from the weeks or months required to retrain conventional AI models. Development productivity has also soared with AI agents like Gemini Code Assist and Gemini CLI — Code Assist alone has delivered a 100% improvement in merge requests and a 50% jump in commits per developer. Challenge: The fight against financial crime is a non-stop race against highly sophisticated, rapidly evolving criminal networks. ComplyAdvantage not only needs the latest AI technologies to remain a step ahead but also meet the high standards for resilience, security, and multicloud operations needed to serve global financial institutions. Solution: To support its growing global footprint and provide a highly resilient platform, ComplyAdvantage adopted a multicloud strategy, building with Google Cloud’s planet-scale infrastructure and integrating Security Command Center to manage incidents and outages. Google’s industry-leading AI has also enabled ComplyAdvantage to enhance its fraud and anti-money laundering risk detections with a new intelligent tool chain that uses Google’s Gemini models to screen public data sources like government websites and news articles for adverse media reports — analyzing eight million articles and processing 1.7 articles per day. ComplyAdvantage also introduced AI agents like Gemini Code Assist and Gemini CLI to bring the power of Gemini directly to developers, so they can innovate faster: Code Assist alone has delivered 100% improvement in merge requests and a 50% jump in commits per developer. Results: With Google Cloud, ComplyAdvantage’s ability to outpace financial crime is stronger than ever. With Google Cloud’s infrastructure and Security Command Center, the platform offers 4x more region coverage and maintains over 99.99% availability for its global customers. The new intelligent tool chain, built around Gemini, not only provides 2x more accurate classification and subclassification in adverse media searches — the team can now use natural language to adapt models for emerging risks in a single day instead of spending weeks or months on retraining. Plus, ComplyAdvantage has seen an up to 85% reduction in alert volumes with Gemini-powered agentic auto-remediation. Key takeaways and highlights from our interview with Mark Watson, Chief Product Data and Technology Officer at ComplyAdvantage: → “We doubled down on Google Security Stack. We use Security Command Center, both on Google and on Amazon. Google's ability to integrate with our automated incident management mechanisms has been very important to running an efficient environment. The platform on Google has been pretty much north of four nines since we started running it.” → “We have been using conventional machine learning models on adverse media. We ingest about 8 million articles a day, probably about 1.6 million articles then have to be processed. We’ve replaced all that with a tool chain we’ve built around Gemini that takes a series of models as a recipe, combines all those things into logical flows that you can run. When we’re classifying adverse media, we’re two times more accurate and also when we’re subclassifying that, we’re also two times more successful.” → “We've rolled Code Assist out now to about 90% of the development team. We've seen a hundred percent improvement in merge requests per developer. We've seen a 50% improvement in commits per developer. The impact of that has been significant on the business and it allows us to bring new approaches in for our customers and to innovate far more quickly.” Google Cloud products used: Gemini, Gemini Code Assist, Gemini CLI, Security Command Center Learn more: → https://complyadvantage.com