Alon Arvatz had already helped build a company someone else wanted to buy. Rapid7 acquired IntSights in 2021 in a transaction announced at approximately $335 million. There was a financial reason to stay. There was also the attraction of having another idea, and Arvatz chose the idea.
He told investor Michael Eisenberg that he left Rapid7 before completing a two-and-a-half-year compensation holdback period. More money had become less persuasive than the chance to make decisions and build something of his own again. It is an awkward development for a retention arrangement when the person being retained starts thinking about the exit.
“I left a lot of money on the table.”
Alon Arvatz, in conversation with Michael Eisenberg, 2024
The next company was PointFive. To understand why its chief executive would make that trade, it helps to go back to a smaller financial puzzle, encountered long before there was acquisition money to leave behind.
Interest, loans, and a useful irritation
At Tel Aviv University, Arvatz studied law and accounting. Basic ideas about interest and loans struck him as things people ought to understand before they arrived at university. He had been through years of schooling. How had such everyday knowledge managed to avoid the timetable?
That irritation became practical work. With partners, he developed the financial-education initiative Kanfei Kesef, also known as Money Wings. They took a curriculum into schools, recruited university volunteers and drew support from faculty members. The university supplied people and opportunities as well as lectures.
Another venture, Cyber School, taught cybersecurity to young people. Guy Nizan became his partner there, before the two worked together at IntSights. A subject that could sound forbidding was turned into courses, workshops and a business. Teaching was already part of his working life.
In a later conversation about his university years, Arvatz put the emphasis on the surroundings of formal study: the people he met, the inspiration and the tools that helped him begin. It is a useful correction to the tidy résumé. A degree occupies one line; the relationships formed around it can fill the next decade.
A company built to look outside
IntSights, founded in 2015 by Arvatz, Nizan and Gal Ben-David, worked in threat intelligence. The problem was what an organization could learn about danger beyond its own systems. Relevant signals existed elsewhere, and security teams needed a way to turn those signals into action.
At IntSights, Arvatz’s product responsibilities put him close to that translation. His professional biography described work across product development, threat research and intelligence operations. He also wrote about cybersecurity for an industry audience. Finding something interesting was one part of the job. Making it useful to someone else was another.
Rapid7’s acquisition brought that work into a larger security business. The buyer described combining external intelligence with its understanding of customers’ internal environments. Its announcement emphasized detection, response and remediation. The transaction was announced on July 19, 2021; the acquisition had closed on July 16.
A company transaction, not Arvatz’s personal proceeds.
For Arvatz, the sale also created a new setting in which to discover a problem. The founders would soon find themselves looking closely at the machinery running their own software.
Three founders, fewer arguments
There is a less photogenic skill involved in building companies: learning to lose an argument without treating it as a small constitutional crisis. Arvatz has described having to learn it.
In conversation with Gali Bloch Liran, he recalled an early belief that being certain about a decision meant he should fight for it. That produced conflict. Over time, trust in his co-founders changed the calculation. Keeping the team working together mattered, and expertise needed to count for something.
The approach they developed gave more weight to the person whose field was involved. They sought agreement, with the CEO deciding when consensus failed. It sounds straightforward on paper. Applying it requires people who are accustomed to taking initiative to accept that initiative is occasionally someone else’s department.
He also acknowledged the unease of beginning again after success. A previous exit can help with recruiting and fundraising while giving the founder a reputation to worry about. At PointFive, Arvatz took the chief executive role alongside Ben-David as chief product officer and Amir Hozez as chief technology officer.
The bill nobody wanted to own
The PointFive idea emerged while cloud operations were being transferred from IntSights into Rapid7. Arvatz, Ben-David and Hozez encountered inefficiencies and found that the available tools did not reach far enough into the problem. They had a complaint with the advantage of being specific.
The company launched publicly in June 2024 after eighteen months of development. Its founding account describes applying habits from cybersecurity research to cloud costs: look continuously for patterns, investigate beneath the obvious recommendations and give engineers findings they can use.
The name carries the ambition. PointFive refers to the goal of helping customers achieve fifty percent cloud cost savings, with half the effort and in half the time. It is a compact business plan to fit on a T-shirt. Those halves describe what the founders aim for; individual customer outcomes depend on the environment.
The organizational problem is as interesting as the technical one. Finance can see an expense. An engineer has to decide what can safely change. Between those two activities sits a queue of competing priorities, questions and work.
Salesforce Ventures, explaining its investment in 2024, described that gap between financial operations and engineering. Customers could receive optimization recommendations without enough context to act confidently. The investor’s account emphasized the team’s experience in security research and its effort to uncover waste below the obvious surface.
There is a recognizable continuity with threat intelligence. A warning has limited practical value if nobody knows what to do next. A cost finding faces the same test. Arvatz’s new market gave an old product question a different invoice.
A book for the rest of the room
Alongside the companies, Arvatz became an author. The English edition of The Battle for Your Computer: Israel and the Growth of the Global Cyber-Security Industry appeared with Wiley in 2023, translated from Hebrew by Eylon Levy.
The book’s contents move from cybercrime and military technology into universities, government, investment and entrepreneurial culture. It treats the industry as something with institutions and a history. The table of contents even makes room for the aspiring entrepreneur. Readers are invited to understand how the business came to exist.
Arvatz had wanted to write a book for years. At the launch, a former colleague reminded him of a workshop where he had written that ambition down. His test of accessibility was closer to home: his wife’s grandfather told him he finally understood what people meant by cyber.
That is a satisfying review for a technical subject. In a university profile, Arvatz also listed reading among his interests, alongside chess, basketball and Torah study with his children. The author belongs to a household as well as an industry.
Now the agents are spending
By June 2026, PointFive had raised a $60 million Series B led by Accel, bringing reported total funding to $96 million. The company had more than 100 employees and operations in Tel Aviv, London and the United States. Arvatz was describing a shift toward the costs of AI.
That shift involved reallocating engineering resources and building a new team. He identified hiring experienced AI researchers as a challenge. The familiar startup obligation had returned in a new form: learn what customers now expect, then change the organization enough to deliver it.
In October, he discussed AI spending with Citi Ventures. His argument centered on examining how systems are constructed and used. An invoice can arrive with a total for tokens while leaving the organization short of an explanation of the work those tokens accomplished.
He described TokenShift running on the machine where a coding agent executes, giving PointFive a view of requests as they are assembled. He also emphasized the controls large financial institutions expect: authorization, budgets, audit records and integration with existing procedures. The ambition is to connect consumption with decisions someone can actually make.
For a founder who first encountered the problem in cloud operations, AI adds another place to investigate. It also makes the question of ownership more awkward. Software can now initiate spending through agents, while the bill still needs a human explanation.
The percentage that came first
Arvatz’s interest in money also runs outside the companies. He has said that giving a tenth of his income began when he was a student earning a modest amount. After the IntSights sale, he set aside a share of his proceeds for donations and returned to institutions that had educated him.
By August 2026, he had joined the board of Keshet, a donor-advised fund in Israel. He described wanting to involve more technology founders in organized philanthropy, with education among his priorities. One challenge he identified was consistency: a startup sale happens once, while the causes people support need resources over time.
The idea brings his career back toward those first university projects. Teaching someone about money, explaining an unfamiliar industry and making a technology bill intelligible all require attention to what another person needs to understand.
His career has produced larger numbers since the early school curriculum. The choices around those numbers remain revealing: when to stay, when to leave, what to build and what to give away. PointFive occupies his working day now. The accounting has always included more than the balance.