In an industry that is only getting more complicated in cloud complexity, a three‑person Israeli startup is trying to rewrite how infrastructure is designed.
AutoPipe, founded in 2024 by Unit 8200 alumni Jonathan Or and Nadav Aharon, has built an AI platform that promises to compress weeks of cloud architecture work into a single day. Its pitch is straightforward: most enterprises now run across multiple clouds, most IT projects fail, and the bottleneck is almost always architecture. AutoPipe believes that with their technology, AI can finally fix that.
Or has been building software since he was 13, and later served in 8200’s engineering and DevOps units. He told The Jerusalem Post that the idea behind AutoPipe came from watching the gap between what AI could already do in software development and what cloud teams still had to do manually.
“With 68% of IT and cloud projects failing, the industry can no longer rely on manual prediction of how infrastructure will behave under real‑time conditions,” Or said.
AutoPipe’s platform uses a set of AI agents to perform work traditionally handled by senior cloud architects. The agents analyze customer requirements, apply architectural logic and best practices, and generate full technical deliverables, such as solution designs, documentation, and infrastructure planning outputs.
Speaking to the Post, Or explained how he saw how quickly AI was transforming coding, yet cloud infrastructure, the foundation every application depends on, remained stuck in slow, error‑prone processes. Designing cloud systems by hand, he explained, is no longer viable when real‑world conditions are unpredictable and the stakes are high.
He pointed to a recent example from the UAE, where a major company’s data center was hit by a rocket during the war with Iran in February. Their entire system collapsed because the architecture relied on a single availability zone with no disaster recovery.
“Only one data center held them up. The whole system was running on this one center, and when it was shut down, it cost the company a significant amount of money and time. Had they used AutoPipe before the war, they would have understood that it was at risk,” he explained.
Aharon, AutoPipe’s VP of R&D, described the problem from the perspective of process and explained to the Post that traditional architecture work requires an engineer to interpret requirements, plan the system, and produce extensive documentation- a cycle that, he says, “can stretch from days to months.”
According to Aharon, architecture is even harder because there are higher requirements and far less room for error. AutoPipe’s system of multiple agents is able to produce complete architectures in hours rather than weeks while preserving the reasoning behind every decision.
AutoPipe has a team of five employees and two advisors, including senior technologists and AI researchers. The company is now building on its earlier pre-seed funding and has recently begun a $1.5 million investment round at a $6 million valuation.
The market they are targeting is customers in the US that work on cloud and infrastructure projects, and the global cloud professional services sector - one that is continuing to grow. The company has already completed dozens of customer projects, secured early-paying users, and raised a pre‑seed round from HCS Capital - all while operating quietly from Tel Aviv and in between reserve stints.
AutoPipe’s founders are now trying to turn their early momentum into a broader commercial push. The company is expanding its customer base beyond Asia‑Pacific, preparing a US rollout, and raising a new funding round to support that growth.