Cohesity pushes out RAG-enhanced Gaia GenAI backup search chatbot

Cohesity has announced a Gen AI chatbot called Gaia that can search through a customer’s backups to find answers to conversational questions.

This builds on its Turing AI/ML gateway initiative which used AWS Bedrock to generate business context-aware answers to questions. Cohesity claims it’s the industry’s first Gen AI-powered conversational search assistant to help businesses transform secondary data into knowledge. That’s because retrieval augmented generation (RAG) adds a company’s backup data in the Cohesity Data Cloud to the Gen AI’s Large Language Model (LLM) inputs.

Sanjay Poonen

Cohesity CEO and President Sanjay Poonen said in an canned statement: “Enterprises are excited to harness the power of generative AI but have faced several challenges gaining insights into secondary data, including backup, archived and vaulted data – because every approach requires re-hydrating the data, and painfully waiting weeks for the data to be available for analytics and insights. Cohesity Gaia dramatically simplifies this process with our patent-pending approach using Retrieval Augmented Generation.”

This means that tasks previously carried out by skilled and expensive data scientists querying data warehouses and the like, with specialized coded programs, can now be done by ordinary managers and employees querying backup data.

Poonen said: “Our approach delivers rapid, insightful results without the drawbacks of more manual and risky approaches. In short, it turns data into knowledge within seconds and minutes.”

Gaia is a Saas offering integrated with Cohesity’s Data Cloud and features:

  • RAG AI conversational search experiences across cloud and hybrid environments. 
  • A fully indexed database of all files, across all workloads, and at all points in time that supports the creation of AI-ready indexes for rapid conversational search and responses. Initially, Cohesity will support Microsoft 365 and OneDrive data and will expand to more workloads over time.
  • All indexed data is immediately available for reading without the need for backups to be reconstructed, allowing the Cohesity Data Cloud to function like a data lake with real-time access to data
  • Granular role-based access controls and zero-trust security principles to protect access to sensitive information and help maintain compliance with various regulatory requirements.

Cohesity says customers can use Gaia:

    • To assess an organization’s level of cyber resilience.
    • To perform financial and compliance audit checks.
    • To answer complex legal questions.
    • To serve as a knowledge base to train new employees.

    It has announced plans with the public cloud big three; Azure, AWS and Google, to bring their LLM services to Gaia.  

    As an example of Gaia use, Cohesity’s Greg Statton from its Office of the CTO, said: “If you notice a rise in costs in a region, typically, you would search for dozens of invoices, review and compare them, and see if you can discover the reason for the cost increases. It could take hours, days, or weeks to resolve. With Cohesity Gaia, you simply ask, ‘Why have costs increased in the region?’, and Cohesity Gaia will pull the relevant data from your stored data, analyze it, and return an answer to your question. It’s that simple.” 

    A spokesperson for Rubrik, which announced its Ruby Gen AI chatbot in November last year, said: “If you don’t have a GenAI product available today, you’re not in the game. We rolled out Rubrik Ruby last year and it’s already available.”

    Gaia will be made generally available on March 15. Get a PDF solution brief here.