AI is everywhere. The current renaissance of artificial intelligence (AI) and machine learning (ML) processes that we are currently developing and improving are truly ubiquitous. But just as AI (and the new world of generative AI) is ubiquitous, we generally need to enrich our applications with new AI services. As our new smart applications, databases, and other services benefit from having AI integrated into their core fabrics, we (perhaps) generally see AI in an upward movement. I believe there is.
Bringing AI down to the data layer is also important. Intelligence needs to be injected into the underlying foundation of IT service layers across data stores, data lakes, data warehouses, data fabrics, and traditional databases.
The “Father” of AI
Cohesity is known for its data management technology focused on information infrastructure supporting modern cloud services. In an effort aimed at infusing, applying and diffusing AI wherever businesses put their data, Cohesity is expanding its partnership with Google Cloud to help organizations unlock the power of generative AI and data. We support. In line with the new partnership, the company announced Cohesity Turing (named after British mathematician and “father” of AI, Alan Turing) as a suite of AI technologies aimed at bringing AI to data security and data management. It is clear that it was named in ) was also detailed.
The combination of our partnership with Google Cloud and the new Cohesity Turing service will enable organizations to leverage their entire data estate through a single, secure workflow across all forms of cloud computing.
This isn’t Cohesity’s first data, AI, or Google rodeo. The company has specifically worked with Google Cloud to design an enterprise-grade solution for application backup and recovery. Furthering this industry convergence, Cohesity Data Cloud services will be more tightly integrated with many cloud services, including Vertex AI, Google Cloud’s machine learning platform.
Ask a question about a specific system
This integration will enable customers to quickly search through exabytes of data, for example, to gain insight into “data patterns” that can help identify data anomalies in systems and applications that can help detect threats. will be Similarly, this technology asks a very specific question: when (not a random example) account Juanita tried to validate a new external customer payment channel, an error message occurred in an e-commerce application. You can also be tasked with finding the answer to the question of why it appears. United States, especially Bulgaria. It can also be used to quickly recover data using contextual search. For example (although I don’t need another random example) a request to recover data from transaction records related to new employee appointments made in Maryland between specific dates.
“To apply generative AI transformatively, enterprises need to be able to easily derive rapid insights from their data using state-of-the-art AI/ML models. I am [engineering] Cohesity CEO and President Sanjay Poonen said: “By using our platform, Cohesity not only offers search with built-in indexing capabilities, but also robust security his protocols that allow customers to maintain control and privacy of their data at all times. increase.”
Google Cloud CEO Thomas Kurian also echoed Poonen’s words, partnering in what he calls a “growing open ecosystem” designed to enable more customers to derive value from data through AI. welcomed the extension of
Beyond the hype?
Christophe Bertrand, practice director at Enterprise Strategy Group (ESG), says Cohesity’s data security and management solutions have long had AI and ML capabilities built into them. However, with the introduction of Cohesity Turing, Cohesity is taking a new approach to integrating customer data with responsible AI and governance, ultimately enabling organizations to build trust in the integration and use of AI and data. He suggests.
“this [whole partnership and technology development] Cohesity’s Poonen told reporters in May 2023, “It was never about an AI cleaning exercise. We built a prototype for this use case and people were just wowed. Primary and below. is known to be [newer] With around 90% of corporate data stored in secondary backup data in data stores, it is very difficult to search for insights and the presence of any level of malicious code or actors. This means that you need to “rehydrate” and manage it over time. Consume a complex process. In many ways, the emergence of generative AI is an opportunity for his Cohesity use case more than many other applications and services. Because you can interrogate your backup assets in ways never before possible. And we are the only company. We are doing it at this level,” he added.
Poonen gives an example of natural language search using a large language model (LLM) for backup data exploration and suggests asking: Please provide a two-page summary of the data breach we suffered in 2005, describing the circumstances at the time and the steps we took to remediate the potential loss.
ESG’s Bertrand agrees that “there’s a lot of hype going on around AI,” but he’s happy enough with what’s happening here. The main reason, he said, is that Cohesity is taking what he calls a “pragmatic approach” to bring AI to more use cases.
A closer look at Cohesity Turing is a collection of AI/ML capabilities and technologies integrated into the company’s own multi-cloud data platform and suite of services. As new AI needs emerge, Cohesity says it will continue to evolve its portfolio of technologies delivered by and leveraging this technology, but all (and this is all today). It is a message that must be aligned with AI announcements) In a way, this enables organizations to use AI responsibly and safely. This use case is all about using AI to gain deeper insight into security risks and extract more value from your data.
Data entropy detection
Examples of AI/ML capabilities found in Cohesity Turing include advanced modeling ransomware anomaly detection and advanced data entropy detection, meaning that data objects, values, records, and stores of any form become chaotic and fragmented over time. , disconnects, and even features to detect overlapping trends. . This entropy detection enables data scientists to “see” anomalies in ingested data, providing early warning of threats hidden in protected data.
You can also find threat intelligence here. Cohesity says it designed the technology to provide a highly curated and curated threat feed. This feed is used in conjunction with ML models trained on millions of samples to accurately detect threats and automatically keep your organization up to date. In addition to data classification capabilities that identify sensitive data, Cohesity also built predictive capacity planning. This is a feature that enables customers to meet their current and future business service level agreements (SLAs) with machine-driven recommendations using the included “what-if simulator”. ‘ Test the system you may be changing.
“Additionally, in the near future, customers will be able to leverage Cohesity Turing to gain even more insights from their data through search augmented generation (RAG) AI model workflows. We believe it will enable us to gain deeper insights and discoveries from our data, as well as quickly find the content of petabytes of data.Cohesity has filed a patent in this area,” Poonen and team said in a technology product statement. said in.
So we are faced with the following question. We incorporate AI into enterprise technology systems at every possible layer and in increasingly fine-grained ways to help control every piece of software, and even the precision-engineered components of hardware. The simple answer is yes, but only if it is unbiased, accountable, accountable, and within the guardrails that define regulatory compliance and governance.
interact with data
Breadth and granularity may be considered in the context of technologies like Cohesity Data Cloud. Cohesity Data Cloud is built on a proprietary distributed file system for multi-cloud (enterprises using multiple cloud service provider (CSP) hyperscalers to run their cloud assets). Designed for massive scale across on-premises, cloud, and edge Internet of Things (IoT) environments. According to the company, users can leverage their platform for his AI to benefit from large-scale language models (LLM) and natural language processing (NLP) while still maintaining control over their data. . This means that you can use everyday language to ask questions about your data. .
“Hello, Data. Can you see who has role-based access control to service disruptions, potential security risks, and retail sales records in Florida?” It may sound like a fantastic idea, but this kind of behavior is already a reality.
You may not be able to talk to plants and flowers yet, but you can now talk to your data.
follow me twitter Or LinkedIn.
