Eon turns cloud backups into SQL-searchable Parquet data lakes

Interview. Israeli startup Eon pulled in $127 million in funding in October 2024 for its cloud backup posture management (CBPM) software which turns cloud backups into assets which can be granularly restored and have searchable database snapshots. In December it raised another $300 million, and said it’s the world’s the fastest growing cloud infrastructure SW company.

We wanted to understand how Eon’s tech differs from that of other SaaS backup providers such as Cohesity, Commvault, Druva, HYCU, and Rubrik, how is it going to use AI and what are its cyber-security inentions. Co-founder and CTO Rom Kimchi answered our questions.

Blocks & Files: How does Eon’s tech differ from that of other SaaS backup providers such as Cohesity, Commvault, Druva, HYCU, and Rubrik. They are making their backups searchable, using AI.

Ron Kimchi: Eon is the only cloud-native backup platform providing a smart, super-efficient storage layer designed for backup data. Traditional, snapshot-based backups are complex and costly for enterprises to manage, and require an all-or-nothing restoration, coupling the backed-up data with the old operating system and end-of-life software that were installed on the cloud resource at the time of the backup. The reality is, backup data just sits there, while enterprises pay 10-30 percent of their cloud spend on backup.

Our approach breaks from an outdated standard. We enable instant access to cloud backups by transforming them into immutable, incremental Parquet-based data lakes. With Eon, customers can run SQL queries directly on their backed-up data, all without any complex restore or ETL processes. We are the only vendor supporting that in the cloud. We optimize cost, simplify data recovery, and improve security. We’ve fundamentally changed how enterprises can view their backups, from a necessary cost center to a viable data source for business initiatives.

Eon is also distinct in its cloud-native and multi-cloud design. The platform allows portability between major cloud providers such as AWS, Azure, and GCP, without the vendor lock-in, complexity, or high costs typically associated with this process. Through automated deduplication and compression, Eon customers see up to 50 percent savings on cloud backup storage costs, while also gaining entirely new use cases for these petabytes of data. This design also ensures that during a single-cloud outage, enterprises retain access to their data, keeping vital workstreams online.

Blocks & Files: How is Eon going to use AI in its technology?

Ron Kimchi: Eon continuously and automatically scans and classifies backup data as it’s ingested using LLMs. The platform classifies and tags unstructured content, detects data anomalies and embeds metadata in vector databases for RAG.

Eon’s platform also empowers enterprises to use their historical data for their own AI initiatives, including training and fine-tuning models. The backup data can also support deeper analytics on archives that were previously inaccessible or too costly to use.

And because AI is only as strong as the data fueling it, unlocking backup data becomes a true force multiplier. By opening up this massive, unused dataset in the enterprise, Eon gives organizations the foundation they need to build AI systems that deliver measurable value in production.

Blocks & Files:How is Eon providing cybersecurity?

Ron Kimchi: Ransomware has shifted aggressively into the cloud, with attackers targeting cloud native workloads, including managed databases, object storage and VMs. Traditional security tools built for on-prem environments struggle to protect data in today’s dynamic, managed multi-cloud architectures.

Eon addresses this gap with the first cloud-native ransomware protection package, purpose-built from the ground up for multi-cloud environments. Unlike traditional tools that were made for on-prem storage and retrofitted for the cloud, Eon provides a true logical air gap protecting your cloud, as well as immutable backups, advanced threat detection for both file systems, self-hosted databases, and managed databases, with granular, file-level or database record-level recovery. It allows for recovery of clean database records and files (without restoring vulnerabilities).

Eon’s approach fully aligns with the NIST Cybersecurity Framework, covering identification, protection, detection, response, and recovery. For enterprises, that means end-to-end cloud ransomware resilience built for the threats they face today, not the environments their legacy tools were designed for