Rajat Bhargava and the question worth asking twice

The first boardroom had a chalkboard. In 1994, Rajat Bhargava and his NetGenesis colleagues met their early investors in an MIT classroom, around a wooden table that Brad Feld later remembered as looking a century older than the company. Feld and Will Herman had each put in $25,000. There were no PowerPoint slides. There was chalk, which at least made revising the business plan a fairly inexpensive operation.

Three decades later, the names still fit together. Bhargava leads JumpCloud. Herman became his co-author on The Startup Playbook. Feld wrote a foreword for a later edition. Their early meeting offers a better opening to Bhargava’s career than a procession of company logos: people taking a young business seriously, then continuing to help one another think.

His work has moved through several generations of technology. The recurring interest is in the practical arrangements underneath them. Companies needed to understand their websites. IT teams needed to connect people to machines and applications. Now software agents need identities and someone answerable for their actions. Each wave brings its own invitation to start again.

A company before the career ladder

In July 1994, MIT News identified Bhargava as net.Genesis’s CEO, a senior and a master’s candidate in electrical engineering and computer science. The student company would help businesses establish an internet presence. Among his colleagues were Matthew Gray, Matthew Cutler, Eric Richard and Craig Wisneski. The web was young enough that helping a company find its place on it was a business proposition in itself.

Bhargava’s early career therefore began with a peculiar reversal. He had a CEO’s responsibilities while still being a student. Before there could be much of a corporate apprenticeship, there were customers to find and decisions to make. The title arrived early; the education in what it meant would take longer.

He later recalled that NetGenesis explored several products before concentrating on website analysis. His interest in the challenge had an earlier expression: he completed high school in three years after splitting his days between a math-and-science magnet school and a regular public school. Having finished the credits and gained admission to MIT, he saw little reason to stay for another year.

The ventures that followed included Service Metrics, Interliant, Quova, StillSecure and Yesware. Looking back on Service Metrics, he described an 18-month run ending in a $280 million acquisition by Exodus. Yet his assessment of a company could also turn toward the people who began careers there and went on to build elsewhere. The financial result and the human afterlife were both part of the accounting.

The idea that waited

JumpCloud’s origin has a pause in it. Larry Middle remembers discussing a next-generation directory service with Bhargava while they were building StillSecure in 2007 and 2008. They wondered why the market seemed to leave so little room for an alternative. They could see something worth investigating, but could not yet work out the answer. The idea waited.

Middle dates their return to it to 2013. That gap is a useful detail in a career often summarized by the number of companies started. Experience did not make every problem immediately soluble. A familiar frustration could remain unfinished business until another attempt became possible.

JumpCloud launched at TechCrunch Disrupt in San Francisco in September 2013. Its early pitch addressed server administration and automation: provisioning, configuring, managing and monitoring. On December 16, the company announced Bhargava’s appointment as CEO and a new Boulder headquarters. It said hundreds of small and midsize enterprise customers had joined in under three months.

That beginning helps explain the work’s appeal. Server maintenance offers few occasions for applause. It does offer plenty of occasions to lose an afternoon. Saving an administrator from repetitive work can have immediate value, even if the product requires a rather long explanation at dinner.

The directory became the center of JumpCloud’s proposition. In a Windows-centered office, the connections between an employee and company resources could be organized around a familiar domain. Cloud services, different operating systems and work from different locations made that arrangement less comfortable. JumpCloud’s approach brings identity, device and access management together.

The vocabulary can obscure the ordinary event underneath it. A person starts work. They need access to some things, and should be kept out of others. Their laptop must meet the organization’s requirements. When they leave, their access needs to end. Making those transitions reliable is the kind of problem that stays important after the launch announcement has disappeared.

A playbook with fingerprints on it

By 2018, Bhargava and Herman had been hearing similar questions from entrepreneurs often enough to try answering them in a book. The Startup Playbook was written from the founder’s side of the table. Its subject was the accumulated practical knowledge people exchange when they have already made some of the mistakes.

Their history gave the partnership a particular texture. Herman had helped finance Bhargava’s first company. Now they were organizing lessons for people attempting their own first businesses. The advice traveled across the table, from an earlier investment to a later collaboration.

“Founder = responsibility.”

Rajat Bhargava, writing on the role of a founder

Bhargava’s short equation carries an extensive job description. In his writing, the founder initially answers for the company’s product, team, successes, failures and possibilities. Other people will eventually take over pieces of that work. At the beginning, a missing department does not make its duties disappear. It generally makes them the founder’s duties.

He is equally interested in the work before there is a company. His essay on finding startup ideas treats invention as a process people can practice. Personal needs and conversations with potential customers can reveal opportunities. Waiting for a cinematic moment of inspiration leaves too much to the cinema.

The practical consequence is to examine the business around the idea. A product needs customers who care, a way to reach them and an economic reason to keep supplying it. Enthusiasm helps a founder persist through that examination. It can also make the awkward questions easy to postpone.

His writing about founder characteristics makes room for curiosity and realism together. Ask why a gap exists. Listen when customers do not buy. Pay attention to what colleagues and investors are telling you. He also argues that success needs a definition shared by the people building the business, rather than one borrowed automatically from the people watching it.

The book’s launch had some cheerful arithmetic of its own. The authors priced the Kindle edition at 99 cents for its first week and pledged books to BUILD, which brings entrepreneurship programs to students in under-resourced communities. They had committed 100 copies and planned to give five more for every hundred sold during launch week.

The teaching continues beyond the book. Techstars co-founder David Cohen has described Bhargava and Herman as mentors. MIT Sandbox lists Bhargava on its Funding Board, which supports student entrepreneurs with guidance and funding recommendations. There is a pleasing continuity between the student who needed a first investor and the alumnus helping students work through their own proposals.

Software gets a personnel file

In 2026, Bhargava’s public writing followed identity into another unfamiliar workplace. AI agents can do work inside a business, use its systems and retain access after the original assignment has faded from memory. His question is who takes responsibility for them.

In May, he argued for tying an agent to a human owner. By August, his Machine Speed newsletter was examining what an agent needs: a registered identity, a defined purpose, limited permissions and a lifecycle that includes retirement. The aim is to make accountability part of deployment, rather than an investigation conducted afterward.

One August essay described a conversation with a chief human resources officer and a chief information officer. An agent had sent a customer a communication that was factually correct but wrong for the context. The software had done something consequential in the space between a technical permission and a human judgment.

Bhargava’s proposal was for HR and IT to share the work of governing agents. IT understands access controls and technical records. HR understands roles, ownership and the processes that surround someone joining or leaving an organization. Applying those habits to software requires a conversation the usual departmental boundaries do little to encourage.

The connection to his earlier work is clear. Technology becomes useful inside an organization through a set of permissions and responsibilities. Change the technology and those arrangements need attention again. The early web, the cloud directory and the autonomous agent each create fresh work around that same practical difficulty.

Leave time for another set

“Maybe just enjoy the journey a little bit more.”

Rajat Bhargava, on advice to his younger self

Outside work, Bhargava enjoys tennis. He describes the right balance between work and personal life as an individual choice, and acknowledges that he likes working a great deal. Asked what he might tell his younger self, he suggested enjoying the journey more. It is a modest adjustment for someone whose career has repeatedly supplied a new challenge.

The classroom table is a good place to leave him. The company changed. The technology changed. Some of the people kept turning up, finding new things to build and better questions to ask. A career can be measured in exits and appointments. Bhargava’s also rewards attention to the unfinished idea, the lasting collaborator and the next conversation.