Israel Duanis and the price of saying no

A seven-figure deal is an awkward time to discover whether your company values mean anything. Israel Duanis says Linx Security turned one down. The project would have taken the company beyond the work it believed it could do well. He discussed it with his co-founder and leadership team, then declined. A year later, he says, the relationship produced a different partnership, one that fit.

There is a pleasing inconvenience in this account. Company values are generally agreeable until they interfere with revenue. Then the wording tends to become wonderfully flexible. Duanis describes a more expensive interpretation: a promise about quality should also govern the work you refuse. The conversation continued even after the proposed deal did not.

His account of hiring offers another test of those values. He described asking candidates what they had taught themselves outside work during the preceding three months. The question gave them an easy place to begin, while letting the interviewer hear whether they enjoyed exploring something unfamiliar. A résumé records the skills someone has already acquired. This question leaves room for the thing they chose to learn when nobody had assigned it.

By October 2026, Duanis was leading a company that had announced $83 million in funding. His route there included military technology, threat prevention, vehicle fleets and delivery software. The common question running through that career is surprisingly domestic: with so many things moving at once, who is keeping track of the keys?

Before the permissions, the traffic

Duanis studied physics, mathematics and computer science at the Hebrew University of Jerusalem, followed by systems engineering at the Technion. His professional background includes the Talpiot program, Unit 8200 and leadership of Check Point's threat-prevention business. These are useful qualifications for encountering a complicated system without expecting it to simplify itself out of courtesy.

Then came a detour into transportation. With Lior Gerenstein, he built Fleetonomy, a business applying software and demand prediction to vehicle fleets. The company worked with automotive names including Toyota, Jaguar Land Rover and Audi. Its practical concern was how to put vehicles to better use as customers increasingly expected transport and deliveries on demand.

In a 2018 interview at LA CoMotion, Duanis described the information operators needed: traffic, weather, flights, hotel occupancy and city events. A fleet's workload could change because something happened far beyond the depot. Better decisions required an understanding of that surrounding world. Counting the cars was only a start; the operator needed to understand what those cars would be needed for.

The appeal of this problem is easy to see. A vehicle waiting in the wrong place can be perfectly serviceable and commercially useless. A busy fleet can still be badly organized. Software earns its keep by connecting facts that an operator would otherwise have to piece together. Years before Linx, Duanis was already explaining why an inventory needed context.

An exit with another job attached

Via announced its acquisition of Fleetonomy in October 2020. The buyer wanted to bring the team's fleet-management expertise into its expansion in logistics and delivery. For Duanis, selling the company led into operating inside a larger one. By July 2021, he was Via's vice president overseeing logistics and last-mile delivery platforms.

The acquisition also joined two founding teams. Via co-founders Daniel Ramot and Oren Shoval publicly praised the Fleetonomy team and its products, describing shared plans for transportation and delivery. Duanis would continue working on the problem he had built his company to address, with a different organization around him.

He appeared on Joe Lynch's Logistics of Logistics podcast to discuss the difficulties of that final stretch to a customer's door. The work covered software for retailers, operators and public organizations, with an emphasis on serving demand efficiently. The customer's expectation was immediate and personal. The system behind it had to reconcile many such expectations at once.

Looking back from Linx, the transportation chapter makes his career more interesting than a straight ascent through security companies. Delivery software has to remain useful when plans meet traffic. Identity software has to remain useful when an employee changes jobs, a contractor finishes work or a new application appears. The products are different. Both ask their builders to take operational mess seriously.

A friendship becomes a founding team

Duanis and Niv Goldenberg founded Linx in 2023. They had met more than two decades earlier during their military service. Goldenberg brought experience from Adallom, Microsoft and Transmit; Duanis brought his security background and the experience of building and selling Fleetonomy. Their partnership had a history considerably longer than the business plan.

That history also connected them to Assaf Rappaport and Yinon Costica, who became Linx investors. Cyberstarts and Index Ventures backed the company, alongside other investors from the security world. When Linx emerged publicly in July 2024, it announced $33 million raised. The new company arrived with established relationships behind it.

A long acquaintance cannot settle every product decision. It does mean the founders begin with more evidence about each other than a promising introductory meeting can provide. Their business would ask enterprises to trust software with sensitive access decisions. Before they could make that request, two people had to decide they could build something together.

The short question with the long answer

Duanis framed Linx's task in an early interview with a question: “who should have access to what at your company?” He also explained the name. Linx maps the links between identities, assets and the organization. The wording is compact enough to fit into a conversation. The answer can occupy an entire enterprise.

Consider an ordinary change of role. A person needs access to new tools, while permissions from the previous job may remain. Multiply that administrative chore across applications, employees and contractors, then add service accounts and automated processes. The organization chart shows responsibility. The access records can tell a rather different story.

Linx's stated approach is to bring identity, security and IT teams into a shared platform. Each team has its own responsibilities and tools; the same account can look different from each desk. Duanis's proposition depends on connecting those views so that the person reviewing a permission can understand the relationships around it.

It is a business built around a neglected kind of drama. Nobody applauds when yesterday's access disappears on schedule. Yet that quiet administrative success is precisely what a security team wants. The tension sits between the company's changing needs and the permissions that have failed to change with them.

Autonomy, with somebody checking the work

By March 2026, Linx had announced a $50 million Series B led by Insight Partners, with Cyberstarts and Index Ventures participating. Duanis described plans to expand development and sales. The company reported more than 60 customers, primarily in North America, and a tenfold increase in revenue over the preceding year. Those were company-reported figures, giving a sense of the scale it was pursuing.

His attention was also moving toward AI agents, which can inherit access once associated with people. A software agent that performs work inside a business needs permissions. Its usefulness and the consequences of its mistakes both depend on what those permissions allow. For Duanis, this extends the identity problem into another population of actors.

In May, writing about Linx Autopilot, he described customers wanting automation together with an inspectable record of decisions. They preferred to begin with recommendations, move to approved actions, and grant more freedom within specific tasks. He discussed narrowly defined agents for reviewing access and identifying unjustified administrative privileges.

“We make autonomy boring.”

Israel Duanis, May 2026

That is an unusually restrained ambition for an AI product. The point is practical: a buyer has to defend what the system did. Duanis's argument makes the record of a decision part of the product's value. The machine gets responsibility in stages, while the people using it retain a way to examine its work.

The hundred people behind the number

On October 6, Duanis marked Linx's hundredth employee with a note about the decisions people make when joining a young company. He treated each hire as an act of trust. A person leaving an established job is making a bet that cannot be reduced to the founder's confidence or the investor list.

He also recalled a prospective customer's security incident during a Linx company event in 2024. In his account, Goldenberg and two teammates joined a call within minutes and delivered a report the next day. The organization was still evaluating Linx. The anecdote gives his discussion of customer commitment a concrete test: help was needed before the contract was secured.

The emphasis on people appears beyond the formal company announcements. Duanis has publicly shared a Linx event welcoming employees' children into the workplace. His older Founders Village biography names family time among his interests, along with playing drums, piano and guitar. Three instruments suggest a reasonable appetite for coordination, although no amount of systems engineering guarantees a decent guitar solo.

There is room for a little irreverence, too. In a LinkedIn post about a blacksmithing workshop for identity-security teams, he complained that the team had organized a particularly appealing event without inviting him. A colleague's reply conceded that he could come next time. Even a chief executive occasionally has to request access.

The seven-figure refusal, the long friendship and the hundred hires all return to decisions about responsibility. Duanis is asking customers to delegate consequential work to software, and asking employees to commit their careers to the company building it. His public accounts offer a standard by which to judge that request: define the work carefully, keep the promises specific, and make sure the people who said yes have reason to do so again.