A week after Port raised $100 million at an $800 million valuation, Zohar Einy sat down with IsraelTech at Google for Startups Tel Aviv to talk through what the company is actually building, why AI agents are making engineering organizations more chaotic before they make them better, and what it will take for Israel to build category-defining companies outside of cybersecurity.
Einy is the co-founder and CEO of Port. He previously co-founded Aporia. He is 31 years old, recently married, and grew up on a moshav with horses and chickens. He runs three to five kilometers on Rothschild Boulevard most mornings. These details matter because he talks about company building the way someone talks about something they cannot imagine doing differently.
Where Port Came From
Port was born from a problem Einy and his co-founder and CTO Yan Koslowski lived inside the Israeli military. Both served in Unit 8200. Koslowski led a DevOps team of 30 to 40 engineers responsible for serving around 2,000 developers. Einy was one of those developers.
The system for getting what you needed as a developer was filing a ticket to Koslowski’s team. That team was talented but overwhelmed. Developers who wanted to move fast were blocked by a queue that existed because there was no better mechanism.
Koslowski built one internally, a system called America that let developers self-serve: request a permission, click a button, get it automatically without waiting weeks. That experience, feeling the problem and seeing it solved once at scale, became the foundation for Port. The insight was that engineering organizations should be self-sufficient. Developers should not be blocked by operational processes. The question was how to build that for every organization, not just one.
What Developers Actually Do All Day
One of the more useful parts of the conversation is Einy’s breakdown of what engineering work actually looks like in practice, because it is not what most people picture.
Writing code is roughly 10 to 15 percent of a developer’s time. The rest is managing incidents when something breaks in production, remediating security vulnerabilities, meeting compliance and privacy standards, and owning the entire lifecycle of the software they ship. This expanded ownership model, what AWS CTO Werner Vogels coined as “you build it, you own it,” emerged from the DevOps era roughly 12 years ago. Before that, developers wrote code and handed it to IT. Now they are responsible for everything end to end.
That is a meaningful expansion of scope, and it is the source of both the productivity gains and the new pressures that Port is built to address.
The Agent Chaos Problem
The challenge Port is solving right now is what Einy calls agent chaos.
Engineering teams across the industry are trying to bring AI into the full development lifecycle. They start with coding assistance, tools like Cursor or GitHub Copilot, and from there they want to connect AI agents to all the systems they use: their cloud infrastructure, GitHub, Datadog, Wiz, and others. They give those agents tasks: resolve this incident, fix these vulnerabilities.
The problem is that agents need context and high-quality data to operate well. When they are connected to thousands of different systems without a clear source of truth, they get confused, make mistakes, and gradually lose the trust of the teams using them. Einy gives a concrete example: an agent tasked with solving a latency problem in production decides the solution is to delete the entire database. Technically it resolves the latency. It also destroys all the customer data.
The issue is not that agents are fundamentally unreliable. It is that there is currently no layer between agents and production systems that enforces guardrails, maintains visibility, and keeps humans appropriately in the loop. Agents are running reckless, in Einy’s words: no human oversight, no audit log, no clear understanding of what is happening.
Port sits in that gap. It takes AI-powered engineering from working demo to production-grade, by providing the context layer, the governance, and the control mechanisms that let organizations trust what their agents are doing and move at speed without catastrophic risk.
Where Engineering Is Headed
The broader shift Einy describes is from manual engineering to autonomous engineering. AI will write code, resolve incidents, and remediate vulnerabilities. The role of engineers is changing from doing the work to directing, overseeing, and approving the work that AI executes.
This does not mean fewer engineers overall. Einy’s view is that organizations running on AI will need more developers, not fewer, because the volume and velocity of output will increase dramatically and someone has to manage it. The developers who will struggle are those who do not adapt. The engineers who thrive will be the ones who can see the full picture, understand the entire development lifecycle, and effectively direct agents across all of it rather than staying focused on a narrow technical specialty.
On Marketing, Go-to-Market, and Israel’s Blind Spot
Einy is candid about something that comes up repeatedly across the Israeli tech ecosystem: product-focused founders systematically underinvest in marketing and go-to-market.
Israel, he says, produces exceptional technology. The engineering talent is world-class. But building a large company requires more than a great product. It requires the ability to explain what you are doing to a wide audience, build a go-to-market motion, and develop intellectual property in distribution and messaging, not just in the product itself.
He describes his own experience at Port as a process of rewiring an engineering brain to think about go-to-market with the same rigor applied to technology. It was not natural. It required deliberate attention.
His view on how Israel moves from startup nation to something bigger: success stories outside of cybersecurity. Founders follow what has been proven to work. The script for building a security company in Israel is well established, the investors are comfortable with it, the ecosystem supports it, and the exits validate it. To pull founders into other markets, those markets need visible wins. Port, in the developer tooling space, is trying to be one of them. He cites JFrog as the one Israeli company that has gotten close in that space, and says Port is going further.
The stated goal is IPO.
How Port Works With Investors
One detail worth noting because it is genuinely unusual: Port removed the table from its board meetings.
Einy operates with full transparency with investors, sharing challenges and failures as openly as wins. The physical setup reflects that. Board members sit in a circle, no table between them, which changes the body language and the dynamic of the conversation. The idea came from a facilitator at an executive offsite who asked whether they wanted to remove the table. Einy said yes immediately.
He describes TLV Partners, which led investment in both Aporia and Port, as the benchmark for how he thinks an investor relationship should work. Two qualities stand out: they give founders genuine authority over their own decisions rather than directing from the board, and they give honest feedback without filtering it. From there, Port has brought in Team8, Bessemer, Excel, and General Atlantic, specifically looking for investors with that same DNA.
His framing: when you take investment, you are getting married to someone you cannot divorce. The quality of those relationships on the difficult days matters as much as the capital on the day it arrives.
About Zohar Einy and Port
Zohar Einy is the co-founder and CEO of Port, an internal developer platform that provides the context layer, governance, and control mechanisms organizations need to run AI agents reliably across the engineering lifecycle. Port raised $100 million at an $800 million valuation, with backing from TLV Partners, Team8, Bessemer, Excel, and General Atlantic. Einy previously co-founded Aporia. He and co-founder and CTO Yan Koslowski met in Unit 8200 of the IDF. Port is headquartered in Tel Aviv and is targeting IPO.
