ExaWizards' RAGOps now supports AWS' generative AI service Amazon Bedrock

ExaWizards Inc. (Minato-ku, Tokyo; Representative Director and President: Makoto Haruta) announces that "RAGOps," an application template for ExaWizards' AI application development environment "exaBase Studio," now supports Amazon Web Services' (AWS) generative-AI application building service "Amazon Bedrock." With this, customers can pick the most suitable model on Amazon Bedrock and quickly develop and use generative-AI applications that securely integrate with their business data. RAGOps becomes available in May 2024.

As generative-AI adoption accelerates, attention is rising on RAG — which lets generative AI work with each customer's own data. At the same time, customers report issues like "results don't meet expectations" and "we worry about hallucinations spreading wrong answers." In response, ExaWizards designed the RAGOps application template, which continuously improves answer quality from RAG on exaBase Studio. exaBase Studio offers application templates that distill our delivery know-how, and RAGOps is one of them. With these templates, you can add or rearrange data management, AI, and software processing — and develop AI applications while running a working prototype.

* RAG: Retrieval Augmented Generation. Application template: a mechanism that deploys onto exaBase Studio as a starting point — used as-is or customized with other software and data.

With AWS' support, ExaWizards designed exaBase Studio (running on AWS) to call Amazon Bedrock directly. This lets AWS users send and receive business data on exaBase Studio with generative AI more securely.

Amazon Bedrock provides API access to generative-AI foundation models from AI21 Labs, Anthropic, Cohere, Meta, Stability AI, Amazon, and others. It supports multimodal data types — images, not just text.

Bedrock support enables flexible selection of effective foundation models, with potential to improve RAGOps accuracy.

RAGOps can run securely by routing data communication between the exaBase Studio platform and Amazon Bedrock through a private network.

Amazon Bedrock secures and privately protects data and generative-AI applications. Data is encrypted both in transit and at rest. With AWS PrivateLink, you can establish a private connection without exposing traffic to the public internet.

Amazon Bedrock is in scope for common compliance frameworks such as ISO, SOC, and CSA STAR Level 2; the U.S. healthcare-IT security and privacy standard HIPAA; and is GDPR-compliant in the EU.

RAGOps has been available as an application template on exaBase Studio since May 2024. The connectivity to Amazon Bedrock's generative-AI foundation models is also available. It is delivered to exaBase Studio licensed customers via the template feature, and through ExaWizards' professional services. Please contact us for pricing and other details.

In collaboration with AWS, ExaWizards aims to bring new value to customers and expand business opportunities.

exaBase Studio is a development environment for building AI software by combining in-house and external AI models, services, and data. The intuitive design / development UI called "Canvas" visualizes processing — making open design and updates feasible not only for engineers but also for executives and field-knowledgeable people inside and outside the company. Built-in semi-automation of cumbersome development steps fundamentally changes how development happens. exaBase Studio supports an Agile-style in-house development approach where you "improve while you touch what works."