GenAI and agentic models are reshaping financial services by enhancing user experience, fraud detection, payments innovation, and compliance processes.
Key trends identified by John Kain of AWS include end-to-end user experience enhancement, data modernization, and embedded financial services.
Real-time payment systems globally, such as India's UPI and Brazil's Pix, are becoming prominent, influencing customer expectations and posing challenges for infrastructure and fraud prevention.
AI is increasingly utilized for fraud detection in instant payments, with techniques like distributed model training and clean room environments to share fraud data.
Agentic AI, particularly AI agents, is gaining traction in serving customers more effectively through tasks like call center automation, compliance, and personalization.
Models like retrieval-augmented generation (RAG) are improving accuracy and guarding against hallucinations in financial decision-making processes.
AWS is addressing cost concerns related to GenAI through custom chips, model distillation, and flexible model options, offering clients choices to optimize performance and cost.
Clients can leverage AWS models like Nova and DeepSeek R1, which provide cost-effective alternatives for tasks like AML while maintaining performance.
Customers are encouraged to experiment with different models and price points to find optimal solutions that balance quality and cost effectiveness.
Kain highlights the importance of flexibility in model selection and cost optimization for financial institutions adopting GenAI technologies.