Generative AI: Automating storage and backup management
We look at how generative artificial intelligence is being deployed to automate monitoring of repetitive tasks and report on enterprise storage and backup, but also the limits to its implementation
Generative AI – including services such as ChatGPT and Google Bard – looks like being the next wave of business automation.
In IT, a growing number of suppliers now have some artificial intelligence (AI) or machine learning (ML) capabilities built into their management tools. Much of the day-to-day work of IT departments can benefit from automation, where machine learning tools can handle vast amounts of data at speeds far greater than a human analyst.
Storage and backup are obvious areas to target because they represent a significant but often repetitive workload for IT teams. Suppliers are starting to examine how generative AI, with its ability to understand and learn from data and report back in natural language can help manage storage and backup.
AI 101
Generative AI and AI based on large language models (LLM) creates human-like responses to questions or prompts. The technology is already used in customer services, where “chatbots” aim to answer questions online and cut workloads for human operators. Such tools can also be used for research, to create marketing documents or even to create illustrations and artwork.
Automation takes care of repetitive tasks and prioritises issues that need to be passed to a person, and that frees up analysts to deal with more strategic tasks.
“The distinguishing advantage of generative AI is understanding context and generating relevant content,” says Kshitij Jain, head of analytics for the UK and Europe at EXL. One of its benefits over conventional AI is “conversational” or natural language reporting.
Generative AI, data and storage management
So far, most suppliers work with a combination of generative AI and more conventional ML models to automate IT tasks. Generative AI is likely to play a role in the reporting side of IT management, as well as assisting IT departments with the specification and configuration of systems.
“Every day, IT teams see hundreds of predefined tasks fail,” says Mark Molyneux, EMEA chief technology officer (CTO) at Cohesity. “A backup administrator’s job is to examine, reschedule and restart any failed jobs. AI can automate any of these processes.”
AI can, he suggests, automate information gathering and find out why a task or process failed. Potentially, as organisations grow more confident in their use of AI, they could allow automated tools to fix problems too.
Cohesity also recently launched a set of tools, Cohesity Turing, which uses AI for ransomware detection and remediation. A similar approach could be extended more broadly to data management.
“AI can massively reduce the toll on IT and security teams by performing many of the important but tedious tasks itself, and provide comprehensive reporting with clear next steps,” says Molyneaux.
At Commvault, senior director of international systems engineering Jason Gerrard sees a similar picture emerging.
“It is entirely possible for AI to be used for tasks like storage configuration, setting backups and checking compliance,” he says. “It is possible for chat tools to streamline reporting back to users, and some companies are already doing this with virtual helpdesks.”
Another role for generative AI is to help with IT planning, says Sam Woodcock, senior director for cloud strategy and enablement at 11:11 Systems.
“AI and ChatGPT type tools will help customers in the future make more informed and strategic decisions when it comes to procuring, architecting and deploying IT solutions,” says Woodcock.
Generative AI could help with peer reviews of systems, pricing and solution data “and be able to leverage natural language models to gain insight and information simply and easily,” he adds.
Generative AI: Handle with caution?
Despite the potential, however, some industry experts remain cautious about the role of generative AI and LLMs in managing critical IT infrastructure.
To date, there are more use cases based around conventional AI and ML, as well as discriminative AI models that can be used to classify data. But, generative AI and public “chatbot” services in particular come with risks.
“Rather than generative AI, traditional ML is increasingly used to provide recommendations based on data categorisation, with the aim of optimising a fleet of data platforms,” says Patrick Smith, EMEA CTO at Pure Storage.
“Reporting on a storage environment must be accurate. Decisions are made based on this information and any inaccuracies could cause reliability issues or even compromise the integrity of data,” he says.
“The future of storage and backup will undoubtedly be driven by AI and ML, but today that’s not going to be using public chatbots,” cautions Barry Cashman, regional vice-president for the UK at Veritas Technologies.
“In fact, we would strongly recommend against any business using generative AI as a component in their data protection strategy because to get any level of value, companies would need to disclose a level of business-critical data to a third party that could well put them in breach of compliance protocols.”
Read more about AI and storage
- AIOps for storage – potential huge benefits, but lots of lock-in: Applying artificial intelligence and machine learning to storage can help to predict bottlenecks, diagnose I/O issues and recommend upgrades, but suppliers’ products are largely limited to their own hardware.
- Generative AI – data privacy, backup and compliance: We look at generative AI and the risks it poses to data privacy for the enterprise, implications for backup, and potentially dangerous impacts on compliance.
