Doron Sitbon and the useful art of being wrong

Doron Sitbon has a useful qualification for running a software company: he has already watched an idea fail. Among his earlier ventures was a mobile service for the Nokia era, something he would later compare to the ride-hailing businesses that followed. It closed. Being early sounds rather grand until you have to close the business.

His working life began at 13. At 22, he started a company providing services to municipalities, including meter control and mail distribution, and later exited it. He then spent roughly a decade in management at NICE Systems. The sequence gives his career a slightly unusual rhythm: entrepreneur, corporate manager, entrepreneur again. A conventional ladder would have required rather less rearranging.

Years later, while advising on a quality management implementation, he watched a project consume about eighteen months and substantial investment. The project finished successfully. Sitbon came away wondering why success required quite so much time. That irritation would become a business question, and eventually a product.

The fifth business had to earn its place

By 2020, Dot Compliance was being described as his fifth business. Sitbon’s explanation for its survival was refreshingly unceremonious: he had tried several ideas, and this one lasted. He described himself as someone drawn to experiments. During the first year of product development, he met more than 300 prospective customers. Enthusiastic agreement was scarce. He kept listening.

“By nature, I like to experiment.”

Doron Sitbon

There is an appealing lack of theatrical secrecy in his approach. He encouraged founders to discuss ideas widely and seek people who would disagree. He also spoke about a circle of fellow founders and CEOs who could help when the job became lonely. An entrepreneur needs someone to call when the presentation has ended and the answer remains unclear.

He extended that permission to experiment to employees, alongside training and room to exercise judgment. The ambition was an enjoyable workplace where people could develop and take pride in their work. That is a more demanding management task than putting a cheerful slogan on the wall. A slogan never has to decide what to do after an experiment goes wrong.

The work between the idea and the invoice

Sitbon’s education was in industrial engineering and information systems at Tel Aviv University. The combination is apt for someone interested in how organizations move work around. A system can be technically impressive while the person using it spends the afternoon waiting for the next approval. Engineering the software and engineering the working day are related jobs.

He also founded DOT Project Management. Its consulting work spans complex projects, and his professional background includes customer support, business analysis, customer relationship management and workforce optimization. Those are fields in which a diagram eventually meets a customer with a deadline. The diagram tends to have the easier afternoon.

Dot Compliance, founded in 2015, took a specific position in enterprise software: a quality management system should arrive with much of the necessary preparation already done. Built on Salesforce, its platform comes with configured processes and validation material. The company now reports more than 500 customers in 30 countries. Sitbon remains its founder and CEO.

The product handles such organizational chores as documents, training, changes, audits and supplier oversight. These are tasks with a persistent appetite for coordination. A document has a version. A person needs training on that version. An approval has an owner. Someone later needs to establish what happened. The ambition to make these processes ready for use gives Sitbon’s implementation complaint a concrete commercial form.

By July 2022, the company was reporting more than 200 customers across over 15 countries and three successive years of triple-digit annual growth. It employed 80 people, 60 in Israel. A $23 million Series B led by Israel Growth Partners, with Vertex Ventures and TPY Capital participating, brought announced funding to $33 million.

The figures marked a change in the scale of Sitbon’s undertaking. A founder who begins with a complaint about implementation eventually inherits a larger implementation problem of his own: making a growing organization deliver consistently. More customers mean more occasions for the original promise to be tested. The next sale is welcome. The next successful deployment is what gives the promise substance.

From keeping records to reading them

In April 2024, Dot Compliance announced a further $17.5 million in a Series B extension, co-led by IGP and Vertex with TPY participating. Announced total funding reached approximately $50 million. Sitbon described the direction in a sentence: “Dot Compliance is transforming into a data company.” The phrasing mattered because it moved the ambition beyond organizing records.

The company was developing Dottie AI to examine text and other data, find relationships and provide insights. It also outlined a vendor-agnostic, hybrid offering. For Sitbon, the opportunity was to make useful analysis available across the information an organization already possessed. That remained a stated direction, with the practical work of connecting systems following behind it.

When discussing AI assistance in 2023, Sitbon reached for an image with considerably more personality than the average enterprise software diagram: an Iron Man suit for a quality manager. The point of the metaphor was additional capability around the professional. The person would still bring judgment; the software would help examine information and handle work.

He emphasized combining technical expertise with knowledge of the customer’s field, and making the result accessible to people without technical training. That is a revealing preference. A complicated tool can advertise how much its makers know. A usable one has to account for what its customer is trying to accomplish before the next meeting.

His 2024 writing developed a related picture of professional work. People who once passed documents around had become coordinators of processes; with richer data, they could increasingly investigate patterns. He argued that organizations would need data literacy and trustworthy information to support that change. Buying an analytical system would leave a considerable educational assignment behind.

He also raised an uncomfortable question about expertise. If software takes over the repetitive work through which people once learned a field, how will the next generation acquire judgment? It is a good question for a repeat founder to ask. Some knowledge arrives through the very friction a product is designed to remove. Saving time and preserving understanding require attention to different parts of the working day.

A calculator close to home

One small episode brings Sitbon’s interest in faster implementation back inside his own company. Dot Compliance had a complicated sales compensation system, connecting commissions and bonuses to business performance. Changing it was cumbersome. The organization used Tribal’s AI development platform to build an application around its Salesforce and NetSuite data.

Dot Compliance reported 20 hours of development on the project. The application gave salespeople a view of commissions, allowed finance to configure plans, and supported payment review and export. The appeal is easy to understand: compensation is a particularly effective way to discover whether the numbers in a business system make sense to the people reading them.

A founder selling easier deployment could hardly ask for a more domestic test. There was no need to invent an exotic problem. His own organization had a familiar one, complete with rules, exceptions and colleagues who wanted a dependable answer. The episode gives his interest in AI-assisted development an ordinary office address.

Even Iron Man needs a supervisor

By 2025, Sitbon was discussing the difficulty of getting AI’s capabilities into the precise place where someone could use them. His description of the last-mile problem emphasized domain knowledge, careful deployment and organizational preparation. He expected more multistage work carried out by agents, which would make supervision more complicated.

His governance writing put responsibility alongside capability. An evolving system can change behavior, lose accuracy or conceal assumptions inside a chain of tools. Sitbon argued for traceable decisions, access controls, clear ownership and continuing review. These are his proposals for managing the technology, rather than a claim that a clever model automatically brings those disciplines with it.

He has a taste for science-fiction references, invoking Asimov as well as comic-book heroes. The contrast is amusing: the imagined machine gets rules at creation; the actual organization must assign people to write, check and maintain them. Somewhere between the superhero suit and the permissions setting sits a manager who needs to know what happened on Tuesday.

In his January 2026 outlook, Sitbon anticipated agents working across different enterprise systems, more robotics and a changing software development process. He reported that roughly 60 percent of the code in recent Dot Compliance releases was AI-generated or AI-assisted. The categories overlap; the figure describes his company’s account of its work, rather than a measure of autonomous programming.

His latest management analogy is quieter. In October 2026, he proposed treating an AI agent like a new hire: give it a defined assignment, introduce it properly, review its work and establish when a person should take over. He also argued that managers need training in supervision. Installing capability leaves someone with a continuing job.

The advice brings this part of his career back to an old concern. A system’s arrival is an event. Making it useful is a relationship between the system, the process and the people involved. Sitbon began by questioning why a completed project took so long. He is now asking who should direct the software after it starts working. The invoice has been paid. There is still plenty to do.