Amazon has recently announced the general availability of Guardrails for Amazon Bedrock, which provides top-notch safety protection on top of the existing capabilities of foundation models. This feature helps customers block up to 85% of harmful content. Guardrails is unique as it is the only solution from a major cloud provider that offers both built-in and custom safeguards in a single offering. It is compatible with all large language models (LLMs) in Amazon Bedrock, as well as fine-tuned models.
Thousands of customers trust Amazon for Gen AI applications
Major industry players such as the New York Stock Exchange, Ryanair, and Netsmart are already benefiting from Amazon Bedrock’s capabilities to drive innovation, enhance productivity, and improve user experiences. The NYSE uses Amazon Bedrock to simplify complex regulations, while Ryanair empowers its crew with quick access to country-specific regulations and manual excerpts.
Salesforce views AI as a crucial component of its commitment to providing personalized experiences across its applications and considers Amazon Bedrock an essential part of its AI model ecosystem.
“As we incorporate generative AI and help customers deliver practical AI on their unified data, we want to assess all available foundation models to ensure they meet our customers’ needs,” explains Kaushal Kurapati, SVP of AI Products at Salesforce. “Amazon Bedrock plays a vital role in our open ecosystem approach to models, and the new model evaluation feature can streamline our model selection process. With criteria ranging from technical aspects to more subjective factors like style and brand relevance, this capability enhances our productivity and simplifies model deployment for our customers.”
Lan Guan, Chief AI Officer at Accenture, a leading global professional services firm, believes that the latest enhancements in Amazon Bedrock will enhance their foundation model services.
“Our clients rely on Accenture for our AI expertise to solve their core business challenges and drive growth at an enterprise scale. The new Amazon Bedrock features, including Guardrails and Model Evaluation, combined with the latest Amazon Titan models, will strengthen our flagship foundation model services and accelerate the development of powerful generative AI applications for our clients. With a wider range of model options, Amazon Bedrock offers clients greater flexibility to meet diverse organizational needs. These new features also enable our clients to implement responsible AI practices, fostering long-term value and building trust.”
