Breaking the cloud backup ‘black box’ with intelligent data mapping and retrieval

Since the earliest days of computing, enterprises have used backups to keep their business-critical information protected. A successfully established cloud backup posture ensures the organization goes untouched during unforeseen events like natural disasters or system failure. However, even after multifold growth in the scale and complexity of enterprise tech stacks, the approach to setting up these backups has largely remained the same: static and error-prone.

To fix this, Eon, an Israel and New York-based startup founded by a team of ex-AWS engineers, has come up with a new cloud-native backup solution that continuously maps and backs up resources for enterprises, depending on the type of data involved. Most importantly, once these backups are ready, it makes them usable by allowing users to retrieve specific files or records according to their needs. 

The year-old company is challenging the status quo in the cloud backup domain, giving enterprises an entirely new outlook on how they can back up their datasets and applications and utilize those backups when in need. According to Gartner, companies are spending $596 billion on cloud infrastructure and planning to increase their investments to over $720 billion in 2025. Out of this, about 10% goes to backup infrastructure, translating into millions per enterprise yearly.

The company was founded a year ago and is already making waves with its patented back storage technology. It has roped in dozens of customers across sectors and raised nearly $200 million in funding, with the latest round of $70 million valuing the company at $1.4 billion. 

Traditional cloud backup doesn’t keep up

Today’s enterprises heavily depend on cloud infrastructure, thanks to its ability to scale quickly and efficiently. Each cloud instance can potentially host a variety of AI and ML applications and petabytes of data, utilizing virtual machines with block storage, object storage, elastic file systems, databases, data lakes and data warehouses.

In this massive, fast-moving ecosystem, creating a cloud backup becomes quite a task. First, one has to cover an endless, rapidly growing wave of cloud assets, from every active application and database to resources that have been shut down or moved. Then, after identifying the resources, they have to manually tag them with metadata labels (key-value pairs for easier organization and filtering) and create snapshots. These are point-in-time backups that can be configured with varying retention periods, allowing users to restore the assets at any given time within the retention window.

Over the years, these snapshots have evolved, providing enterprises with capabilities like automation (after initial configuration) and encryption. However, on a granular level, they haven’t kept up with enterprise needs and their cloud environments that spiraled out of control. Resources still have to be tagged on a massive scale, which can be time-consuming and error-prone, and the cloud backup has to be restored in the form of full servers/volumes, even when the need is just to recover a few specific files. 

Ofir Ehrlich, who founded CloudEndure with Gonen Stein and Ron Kimchi and later sold it to Amazon to bolster AWS’ disaster recovery and cloud migration efforts, saw this problem firsthand when working with the company’s large enterprise customers. 

“The challenge with these traditional snapshots is that they are costly, and act like black boxes, making it difficult to locate specific files or database records, search through them, and retrieve precise data. This complexity leads to higher costs, operational inefficiencies, and slower recovery times—issues that become critical during emergencies,” Ehrlich told VentureBeat.

As a result, the trio left AWS and launched Eon in January 2024 with a mission to provide enterprises with smarter, more capable backups.

How exactly Eon stand out from the crowd?

While most other companies providing cloud backup solutions, including Rubrik, Cohesity, Commvault and AWS Backup, continue to rely on traditional snapshots, Eon took a different approach by developing its own Eon Snapshots, based on cloud-based storage that is optimized specifically for backups. Ehrlich says these snapshots make backup data instantly accessible and searchable, allowing users to locate specific files or database records, search through them, and retrieve precise information.  

At the core, Eon creates snapshots by automating resource mapping, classification and policy association. It keeps on scanning cloud resources on an ongoing basis, automatically mapping and classifying them based on environment type—whether production, development, staging, or QA—and data sensitivity, such as personally identifiable, health or financial information. Once the resources are mapped, it applies customized backup retention policies in line with the enterprise’s specific business and compliance requirements. This ensures the data is not under or over-backed up.

After creating the snapshots, the company makes them accessible to users, with global search across all the backed-up data, granular restoration and seamless recovery between multiple cloud providers (AWS, Azure, GCP). This gives users the ability to not only locate and restore relevant data – down to specific files – but also run SQL queries directly on the backed-up databases to retrieve specific records from tables or spreadsheets. 

This smart restoration ability saves the hassle of restoring/provisioning full databases and can even be used by teams for running analytics, forensics and audits right on top of their backups. The snapshots use machine learning under the hood to remember which cloud application or system the data is connected to, thereby ensuring the connection is maintained as and when the information is restored/retrieved. 

This also prevents misconfiguration or unauthorized access to data.

“Eon stands out in its ability to deliver instant access to specific data. The solution provides precise, file-level restorations across multiple cloud providers without the need to restore entire servers or databases, enabling enterprises to find individual files and database records in seconds,” Ehrlich noted.  

With this approach of automated tagging and instant retrieval, Eon is essentially making cloud backups smart and immediately usable — unlike what’s been the story so far. It even helps reduce costs and operational burdens by excluding unnecessary resources (during mapping) and shortening data retention periods per business requirements, achieved through context-aware backup policy management.

“This work allows us to deliver a better product that replaces traditional, expensive, and black-boxed snapshots, with an alternative that is more cost-effective, autonomously managed and easily usable through instance access,” the CEO added. He did not share the specific cost benefits involved but noted that the company ​​only charges for backed-up storage on a pay-as-you-go basis, with no fees for real-time backup or instant access/lookups.

As of now, Eon is working to scale its offering and actively engaging with dozens of companies across industries such as travel and hospitality, media and entertainment, food services, gaming, financial services, insurance companies, and retailers. Ehrlich noted that they already have numerous deployments in place, without sharing specific customer details. 

However, it will be interesting to see how the company continues to differentiate in the highly competitive cloud backup space. Currently, its smart search and retrieval ability is the differentiating factor (with dozens of patents for cloud storage and data management technologies already filed), but other backup solutions are also beginning to catch up. Commvault, for example, has already started offering capabilities like automated tagging and dependency-based restores.