Eran Shir and the usefulness of a detour

In 2006, Eran Shir proposed a pleasingly specific use for the internet: combine your favorite band’s tour dates with a campsite reservation website and organize a musical camping holiday. This was an example in his introduction to Dapper, the startup he cofounded with Jon Aizen. It was also a rather better sales pitch than “semantic web.” You could explain the technology for twenty minutes, or you could offer someone a tent near a concert.

Dapper would let people take information from websites and put it to work elsewhere. A page could become a feed, a map, an alert, or part of another application. The reader acquired something closer to a workbench. Shir even explained the name: Data Mapper, with the additional ambition of making the web better dressed.

Across the next twenty years, the material changed. Internet routes gave way to web pages, then road imagery, then software built with AI. The useful thread through Shir’s career is his interest in what becomes possible when scattered information can be measured, connected, and used. The career itself has followed a less tidy route, including research, acquisitions, side projects, and two occasions when he handed another person the CEO’s chair.

A library book, then a network

Shir’s interest in physics began early. In fourth grade, he found a library book about the Manhattan Project. He later described school years spent reading, studying, playing music, and doing physics. His academic path included bachelor’s and master’s degrees at the Technion. He also began doctoral work at Tel Aviv University. Commercial ventures would repeatedly compete with the dissertation for his attention.

His early company, Cogniview, explored machine learning and personalized search. The premise recognized that people asking similar questions might want quite different kinds of answers. After that venture, Shir returned to research, working on the structure of the internet. Here was an enormous connected system whose everyday usefulness did not mean its shape was fully understood.

With Yuval Shavitt, he coauthored the 2005 paper DIMES: Let the Internet Measure Itself. The project recruited volunteers to run lightweight measurement software on their computers. Many vantage points could reveal connections that a smaller number of centrally managed machines would miss. Participation was part of the instrument.

By the time of an August 2005 Princeton talk, the project described more than 4,300 installations in 75 countries, producing over five million measurements a day. The researchers wanted to observe how the network evolved, down to routers and up to larger systems. An ordinary PC could contribute a small piece of an unusually large map.

A camping holiday for the semantic web

Dapper brought a related problem into the browser. Information might be perfectly visible to a human while remaining awkward for software to reuse. A concert listing was readable; combining it with accommodation availability was another matter. Shir’s launch post treated that inconvenience as an invitation to make tools.

Users could identify content through a visual process and create a programming interface for a website. Developers could work with the extracted data; other users could try the service’s creation tools. The examples ranged from RSS feeds and Google Gadgets to alerts and mashups. All were ways of giving an existing page another job.

The idea carried a particular flavor of the mid-2000s web, when people were assembling services from other services and giving the results cheerful names. Dapper called its creations Dapps and Dapplications. Beneath the wordplay was a practical ambition: reduce the work between spotting something useful and building with it.

The commercial application developed toward advertising. Dapper’s technology could extract and structure content, then use it in dynamic display ads. Product information from websites could help populate advertising creative. The same underlying ability to understand a page had found a customer with a budget.

“Your creativity is the only limit.”

Eran Shir / introducing Dapper, 2006

The founder changes chairs

In February 2010, Dapper appointed James Beriker president and CEO. Shir, previously president and CEO, became chief technical officer. He continued to oversee product and engineering. The change matters because an acquisition summary can make the preceding years look deceptively simple: founder starts company, large company buys it, curtain falls. Inside the business, responsibilities were already being rearranged.

Yahoo acquired Dapper that October. Contemporary reporting put the deal at roughly $50 million, although the transaction terms were not publicly confirmed. Dapper had raised about $10 million and employed 35 people. Shir said the operation would remain in Israel, with room needed for the growth he expected.

He also supplied the moment with a joke about unfinished academic business: “My professor is still waiting for me to finish.” An acquisition evidently does little to shorten a supervisor’s memory.

At Yahoo, Shir led the global Creative Innovation Center, working on advertising technology and personalization. He also worked with Bruno Fernandez-Ruiz, who would become his next cofounder. Their partnership began inside a large organization, where ideas had to survive the organization as well as the engineering.

Good Karma, with a points system

Shir subsequently became an entrepreneur in residence at Aleph, the Tel Aviv venture firm. One project from that period deserves a place beside the larger companies. In July 2014, he introduced Karma, an app through which entrepreneurs could ask one another for help, offer assistance, and share knowledge.

The design acknowledged a familiar nuisance: the person who knows the answer may be only a few introductions away, while remaining effectively invisible. Karma added points for helping and making referrals. Its leaderboard recognized useful contributors. Even goodwill, apparently, could benefit from a reasonably organized interface.

The experiment had an ending. In 2018, Aleph announced it would close the app in its existing form, after 4,514 questions and 23,856 answers. Eden Shochat credited Shir as its originating figure and described a community that had developed more ways to meet and help offline. A project could do useful work without becoming permanent infrastructure.

By then, Aleph described a wider ecosystem of founder communities, workshops, and specialist groups. Karma had also brought people together at a hackathon and other gatherings. Its closure makes the episode more interesting than a launch announcement alone. The app was one way to help people find one another; the relationships could continue through other arrangements.

The road becomes a research instrument

Nexar, cofounded with Fernandez-Ruiz in 2015, moved the observation point behind a windshield. At its unveiling, the company announced $4 million in financing led by Aleph. Its first proposition used a mobile application to make ordinary cars more aware of what was happening on the road.

The original product had ordinary constraints. Running a camera used battery power. The view was limited to what lay ahead. Using the phone for the application could complicate running navigation at the same time. A network might sound elegant in a presentation; it still had to work on a driver’s phone.

Shir’s aim extended beyond the individual vehicle. Observations from connected participants could help identify dangerous situations and make information useful to others. The company initially launched in Los Angeles and Tel Aviv. Over time, road imagery became material for mapping, infrastructure information, and AI development. The windshield offered a recurring view of a world that kept changing.

A company can outgrow its seating plan

In September 2024, Greenberger succeeded Shir as Nexar CEO. Shir became Chief Product Officer; Fernandez-Ruiz remained CTO. Greenberger brought commercial experience from Lyft and Tesla. The announcement described a company preparing for new demand around AI, with its founders continuing in technical and product leadership.

For Shir, choosing a collaborator had already been a deliberate exercise. In a 2023 talk, he recalled arranging an early Aleph meeting with Fernandez-Ruiz, his former Yahoo boss. Shir already intended to start a company with him, he said, before Fernandez-Ruiz knew it. Previous shared work had given him evidence of how they handled both good days and difficult ones.

On Invested in February 2026, Shir explained the CEO decision in practical terms. He wanted stronger commercial leadership, room to concentrate on AI, and to remain in Israel. His concern about the transition included longtime colleagues adapting to leadership based in New York. Changing the reporting structure still meant changing people’s working lives.

Two new names on the door

By autumn 2026, the road-data chapter had another name. Nexar and Nauto combined, and Empiric Earth launched on September 15. Its leadership page lists Shir as Chief AI Officer, directing AI research and the models supported by its road record. Fernandez-Ruiz is CTO; Greenberger leads the company.

There is also Mithran, founded in 2026 by Shir and Fernandez-Ruiz, with Shir as CEO. It addresses the work inside organizations: helping employees build software with AI, connected to the systems their companies use. The venture’s account of its origins places it in their experience making Nexar itself operate with AI tools.

A useful tool needs somewhere to run, permission to reach the right information, and a way for people to assess what it does. Mithran’s proposition includes that organizational machinery. The gap between a promising demonstration and a working application is familiar territory for founders who have moved through research, web tools, advertising, and cameras.

Return to the proposed camping holiday. Its charm was its specificity: one person, two sets of information, a trip that would be easier to arrange if the pieces could talk. Shir’s projects have grown and changed, but that early example still gives his career a human scale. Somewhere, someone can see the thing they need. The work is making it usable.