MetalBear, a Tel Aviv-based software company whose founders created the open source Kubernetes development tool "mirrord," is today announcing a $12.5 million seed funding round — showing strong support for its vision to speed-up software development cycles at enterprises, reducing the time it takes to write, test and deploy code into production by 98%.
TLV Partners led the round with participation from TQ Ventures, MTF, and Netz Capital, alongside angel investors David Cramer, co-founder of Sentry, and Ben Sigelman, co-creator of OpenTelemetry.
Through mirrord, MetalBear enables developers to run local code directly against their live cloud environment without deploying. The tool intercepts system-level input and output from the local process and proxies it to the cluster, letting developers interact with real databases, application programming interfaces (APIs), and message queues in real time.
This is not a simulation: the code genuinely touches the cloud. To make this safe, mirrord provides guardrails such as traffic routing, queue splitting, and configuration options that limit or isolate side effects. For enterprises, the result is faster development cycles, reduced infrastructure costs, and safer testing in production-like conditions without needing expensive per-developer environments.
The raise coincides with MetalBear being named a Gartner Cool Vendor in Container Management, marking another milestone as enterprises seek ways to streamline development in an era when artificial intelligence (AI) accelerates code creation but testing remains a drag on productivity.
The Testing Bottleneck in Modern Development
Microservices architectures have changed how software is built, but they also make testing more complex. Developers working on individual services struggle to validate their work in the context of an entire application. Some companies try to solve this by offering dedicated cloud environments per developer, but those carry high costs and still require lengthy deployment cycles.
Eyal Bukchin, MetalBear’s chief technology officer (CTO) and co-founder, described this as one of the biggest inefficiencies in software today in a video call interview with VentureBeat last week.
“With microservice architectures, it’s very difficult to run the one or two services you’re working on in the context of the larger application,” he said. “Enterprises can have hundreds or thousands of services, and running the whole thing locally is often impossible.”
In many organizations, local tests rely on mocks that don’t reflect real-world conditions, while shared staging environments introduce queues and delays. “Staging is the best testing environment an organization has, yet it’s often a bottleneck and paradoxically under-utilized,” Bukchin added.
He also pointed out that AI compounds the issue. “AI is accelerating code generation, reviews, and unit tests, but integration testing isn’t solved by AI. The queue just gets longer because all that new code still has to be tested in context.”
How mirrord Works
mirrord allows developers to run and test their local code as if it were already deployed in the cloud. With a single action in the developer’s integrated development environment (IDE) or command-line interface (CLI), mirrord proxies system-level input and output to the remote environment. This enables the local code to interact directly with databases, message queues, and APIs without a deployment step.
According to chief executive officer (CEO) and co-founder Aviram Hassan, who also spoke with VentureBeat on the same video call, the approach doesn’t remove important compliance-driven steps like SOC 2 or ISO standards, but it makes developers far more efficient.
Typically, software developers must undertake a "long loop," a full continuous integration and deployment (CI/CD) process, where changes move through pull requests, automated tests, builds, staging deployments, and production. With mirrord, developers can first validate their code locally against real cloud conditions, then push it through the long loop only once they’re confident it works.
“We let engineers isolate their change in a running ‘plane,’ see how it behaves, and only then go through the full verification. You still do the steps; you just move the long loop later,” he said.
Adoption and Reported Impact
MetalBear says thousands of developers across enterprises such as Nvidia, Amazon Web Services (AWS), and Apple use mirrord. Reported outcomes include up to 80% faster test iterations, a 30% reduction in production bugs, and, in some cases, development cycle time reductions of as much as 98% (from 15 minutes to 10 seconds, in the case of customer CoLab)!
The open source version is available to all, while a paid enterprise edition allows multiple developers to share a single staging environment without interfering with each other’s work.
Features like traffic routing and queue splitting make it possible to scale development teams without scaling infrastructure costs.
“The open source gives developers that local-in-the-cloud loop,” Bukchin said. “The paid version adds what enterprises need at scale on shared environments—guardrails, permissions, and controls to prevent clashes and accidents.”
While the problem becomes more pronounced in large enterprises with complex microservices and staging environments, mirrord has been adopted by both independent developers and Fortune 100 companies alike.
Investors’ Perspective
For TLV Partners, the appeal lies in mirrord’s ability to address bottlenecks that appear only after AI accelerates coding itself.
Partner Brian Sack said most productivity solutions focus on generating code faster, but the real challenge becomes testing and integration at scale.
In his view, mirrord’s design directly addresses this emerging problem.
Founding Team and Expertise
MetalBear’s founders, Hassan and Bukchin, previously held leadership roles in cybersecurity at BioCatch. Their background in system-level engineering enabled them to build a solution that works with complex enterprise environments such as air-gapped clusters and service meshes without requiring major setup or cluster modifications.
The company operates with 25 employees across 14 countries and is serving customers ranging from Fortune 100 enterprises to fast-scaling technology firms.
“Currently it’s limited to apps running on Kubernetes — any flavor, any cloud or on-prem,” Bukchin said. “Conceptually it applies more broadly, but Kubernetes is a huge market, so we’re focused there.”
What It Means for Enterprise Decision Makers
The benefits of mirrord extend well beyond individual developers, offering value to several types of enterprise leaders responsible for ensuring that software is delivered quickly, securely, and reliably.
Engineering leaders who oversee the lifecycle of large language models (LLMs) and other AI applications can use mirrord to shorten the gap between rapid code generation and practical testing.
These leaders often face pressure to move fast without sacrificing performance. By removing deployment delays, mirrord helps them validate new AI-driven features in real-world conditions while keeping projects on schedule.
mirrord further supports AI orchestration specialists by linking local development directly to shared cloud infrastructure, reducing friction and making orchestration tasks less cumbersome.
For data engineering leaders, who build and maintain pipelines that ensure data quality and integrity, mirrord offers a way to cut back on the overhead of supporting multiple parallel environments.
Instead of provisioning costly, redundant setups for every developer, teams can work efficiently against a shared resource pool, allowing these leaders to stretch budgets without compromising reliability.
And for security leaders, mirrord integrates into development without requiring root access or invasive system changes, upholding strong security policies while enabling developers to work faster, avoiding the usual trade-off between protection and productivity.
Taken together, mirrord provides each of these enterprise decision makers with a tool that reduces delays, cuts costs, and makes testing both faster and safer — aligning the needs of technical leadership with the realities of modern software development.
Looking Ahead
With fresh capital, a growing user base, and industry recognition, MetalBear is positioning mirrord as a core enabler of faster, more reliable software development. By bridging the gap between local coding and cloud environments, the company is targeting one of the last remaining chokepoints in modern development workflows.