In his keynote address, Swami Sivasubramanian, Vice President of Database, Analytics, and Machine Learning at Amazon Web Services (AWS), expressed his belief that AWS services and capabilities will democratize the utilization of generative artificial intelligence (generative AI). He emphasized the potential of generative AI to benefit customers across various industries and departments, including engineering, marketing, customer service, finance, and sales. Sivasubramanian stated that generative AI, powered by ultra-large models such as large language models (LLMs), has captured our imagination and reached a tipping point. It will enhance customer experiences by enabling interactions with virtual assistants, intelligent customer contact centers, and personalized shopping services. For employees, generative AI can boost productivity through conversational search, text summarization, and code generation tools. It can also improve business operations with intelligent document processing and quality controls. Moreover, generative AI empowers customers to enhance the production of diverse creative content. To unlock the value of generative AI on AWS, Sivasubramanian highlighted the importance of model choice and customization with customer data. Different organizations will require distinct models for specific tasks, and they should have the ability to securely adapt those models according to their needs. User-friendly tools and purpose-built ML infrastructure are essential for democratizing AI within organizations and enabling low-cost, low-latency responses. AWS offers a service called Amazon Bedrock, allowing organizations of any size and industry to easily build and scale their own generative AI applications.

It provides customers with access to a wide range of foundation models (FMs) through a simple API, facilitating customization using existing data stores. Amazon's extensive experience in AI and ML technology, combined with recent innovations, has made the capabilities of generative AI possible. Amazon Bedrock continues to expand its FM offerings by collaborating with FM providers like Cohere, Anthropic, and Stability AI. These additions bring text generation models, multilingual text understanding models, and improved image and composition detail, offering customers a broad and deep set of resources for their ML journey. Developers can rely on AI agents for Amazon Bedrock to simplify the programming of generative AI applications. Additionally, Amazon OpenSearch now supports serverless vector embeddings, enabling machines to understand relationships across various types of content. Amazon QuickSight, a business intelligence service, has integrated ML innovations and LLM capabilities from Amazon Bedrock. This combination, known as QuickSight Generative AI, enables easier data collaboration and accelerates data-driven decision-making through natural language prompts. In the healthcare industry, Amazon Bedrock empowers health care software vendors with its HIPAA-eligible service called HealthScribe. This service utilizes speech recognition and generative AI to create transcripts and clinical notes, offering relief to clinicians burdened by updating electronic health records. AWS offers instances powered by NVIDIA H100 Tensor Core GPUs, optimized for training LLMs and developing generative AI applications. These instances are well-suited for complex tasks such as question answering, code generation, video and image generation, speech recognition, and more. Recognizing the growing demand for big data, cloud computing, and AI, AWS has released training courses to prepare individuals for AI and ML jobs. These courses cater to developers using Amazon CodeWhisperer, engineers and data scientists working with FMs, executives seeking to understand generative AI's potential, and AWS Partners supporting their customers in harnessing generative AI's capabilities.