Dreeze Exits Stealth With $33M Seed for Cloud Application Recovery

Dreeze is coming out of stealth with $33 million in seed funding to solve a problem most backup vendors don't touch: restored data without a working application. The round is led by Norwest and Bessemer Venture Partners, with participation from Claltech, and it lands alongside the launch of the company's Application Revival platform.

The Tel Aviv-based company was formed in 2025 by Ziv Kedem, Oded Kedem, and Avi Raichel — veterans of the backup and recovery industry with stints at Zerto and Kashya. Dreeze says it already has 70 employees across Israel and the United States. The fresh capital will go toward accelerating product development, expanding sales and marketing, and reaching more enterprise customers.

What's actually broken in cloud recovery?

The problem Dreeze targets sits between a successful backup restore and a functioning service. A cloud application can depend on infrastructure, configuration settings, identities, secrets, encryption keys, and connections to other resources. Bringing back its database addresses only part of that system.

If the surrounding components are unavailable or restored in the wrong sequence, IT teams may still need to reconstruct dependencies manually. Dreeze calls that period the "recovery scramble" — when the data has returned but the application remains unusable.

"Companies have spent years getting better at backing up data, but getting the data back doesn't mean the business is back," said Ziv Kedem, Dreeze's co-founder and CEO, in the announcement.

That reframing matters for how buyers set recovery objectives. A restored storage volume is a milestone, but it doesn't establish that authentication works, required resources are connected, or the application can complete its normal tasks. The endpoint that counts for the business is a service employees or customers can actually use again.

How the Application Revival platform works

Dreeze organizes the platform around four capabilities:

• Know. Autonomous dependency mapping connects resources to the applications they support and generates a recovery plan from the live environment, keeping the plan aligned with changes rather than relying on an aging runbook.
• Protect. Vaulted backup protects data together with configuration, keys, and other application dependencies, using an immutable, air-gapped vault with optional cross-region replication.
• Orchestrate. Revival Automation executes the recovery plan, rebuilding data, infrastructure, configuration, identity, and connections in dependency order.
• Validate. An automated end-to-end test rebuilds a working application — without production impact or manual cleanup afterward, according to the company.

Orchestration is where the sequencing logic pays off. Restore a service before the resources it requires are available, and teams end up troubleshooting a partially rebuilt environment. A dependency-aware plan aims to cut that manual sequencing work.

Testing carries the commercial weight here. A documented plan describes what should happen; a successful rehearsal shows what actually happens in the tested environment. Worth noting for due diligence: the announcement doesn't disclose independent benchmarks, recovery-time measurements, or a comprehensive support matrix.

What does this change for buyers?

Availability is split by cloud, and that distinction matters for multi-cloud deployments. Dreeze says its AWS offering is generally available today. Its Microsoft Azure product is in Early Access, with general availability planned for early 2027.

Jeremy Miller, founder and CEO of Coravant, said in the release that customers' concern is recovering an entire application with its dependencies in the correct order. That's the operational gap Dreeze intends to close: preparing the complete route back before an outage exposes the missing pieces.

The market context favors the pitch. As applications accumulate services, permissions, and infrastructure dependencies, recovery becomes a coordination problem, not just a storage problem. AI adds another wrinkle — automated agents can change environments without adequate preparation for restoration, a risk area Unite.AI's coverage has flagged. Dreeze's announcement addresses the broader question of rebuilding applications after disruption; it doesn't claim the platform prevents unsafe agent actions.

The seed funding gives Dreeze the runway to turn an application-level recovery pitch into a wider enterprise business. Its next challenge is proving that dependency discovery, protection, orchestration, and testing work together across customers' real environments. For buyers, the outcome on offer is concrete: a tested route back to a usable application when a critical service goes down.