Cohesity tuning Turing data cloud front end into context-aware answer generator

Cohesity is extending its Turing AI/ML gateway by working with AWS Bedrock to have business context-aware answers generated in response to user requests and questions.

Back in May Cohesity revealed its Turing initiative in the context of a Google Cloud partnership focused on generative AI and Google’s Vertex AI platform for building and scaling machine learning models. Turing then did not yet have generative AI capabilities and provided ransomware anomaly detection, threat intelligence, data classification and predictive capacity planning. Cohesity was going to add RAG (Retrieval Augmented Generation) AI Model workflows to Turing to help customers get deeper insights and discovery from data or find content in petabytes of data faster. Now it’s working with AWS to do this.

Cohesity CEO Sanjay Poonen said: ”We now see a rapidly increasing demand for AI-powered data insights from customers. With our expanded relationship with AWS, we aim to unleash the power of enterprise AI while leading the industry in leveraging AI-powered data security and management to strengthen the cyber resilience of our customers.”

Turing itself is presented as a gateway to access large language models (LLMs). AWS Bedrock is a fully managed, serverless service that makes foundation models from various AI companies accessible via an API to build and scale generative AI applications. 

Cohesity says Turing’s features will include:

  • Ability to use an index of a Cohesity customer’s Data Cloud contents data to answer questions about business and application data. Cohesity talks about informed and secure answers and deep insights.
  • Customers will be able to query their own data with intuitive, natural-language conversations and have Turing generate simplified insights from it.
  • Turing will provide grounded responses to queries through RAG (Retrieval Augmented Generation).

A Cohesity blog says RAG is used to focus a large language model’s analytical processing on a business’s own data to provide responses tuned to the business, not so general in scope, and with governance.

In our view Turing’s new capabilities are effectively Cohesity’s data cloud co-pilot. It joins Commvault’s Airlie, Druva’s Dru, and Rubrik’s Ruby in providing a conversational virtual white-glove concierge helper to help users navigate complex data protection universes faster.

Cohesity has also worked with AWS and VMware to add Cohesity’s DataProtect Backup-as-a-Service (BaaS) support for VMware Cloud on AWS to provide comprehensive protection and faster recovery at lower costs.

Cohesity will showcase new capabilities for protecting AWS workloads at AWS re:Invent 2023, November 27 – December 1, in Las Vegas. It will show Cohesity DataProtect delivered-as-a-service, including general availability support for Simple Storage Service (Amazon S3), and enhanced support for Amazon RDS (Relational Database Service). This expansion complements Cohesity’s data protection for Amazon EC2 (Elastic Compute Cloud).

Cohesity RAG should be available in approximately six months, and customers can sign up or the Cohesity Turing Early Access Program here.