Tido Carriero and the problem after the product

Tido Carriero had helped make software that salespeople could sell. Then he founded a company and had to become one himself. The résumé offered little protection. Years spent inside Facebook, Dropbox and Segment had taught him how to build products and teams. Asking somebody to buy his own product required a different kind of confidence.

His way through was disarmingly practical: approach the conversation as customer support. Understand the problem. Offer useful advice. Accept that another product might fit better. For an engineer accustomed to diagnosing what was broken, this made the encounter familiar. A sales call could be a place to solve something rather than perform a convincing impersonation of a salesperson.

That adjustment is a useful entrance to Carriero's career. Again and again, his work has taken him to the boundary between a product people enjoy using and a product an organization can adopt. At that boundary sit permissions, security, recruiting, pricing and the awkward question of what the customer is actually paying for. Glamour has been known to miss these meetings.

The friend who taught him to code

Carriero graduated from Harvard in 2008 with a degree in computer science. He joined Facebook when its engineering organization was still small enough to feel like a startup. His early work included Facebook Ads; he later became an engineering manager on the Pages team. Behind a consumer service was a business with customers, operational problems and bills to pay.

In 2012, he moved to Dropbox. He also recruited Alex Allain, the friend who had taught him to code. Allain moved from Boston to San Francisco to join him. There is a pleasing reversal in this: the student had found a project and was now trying to persuade the teacher to come along.

They worked on Dropbox for Business with the early sales team. Peter Ahn, one of those sellers, helped them understand the features required to turn a team product into something larger organizations could buy. Carriero would later credit Ahn with helping him find a comfortable voice in sales. Their working relationship began with a simpler exchange: what does the customer need, and what must engineering build?

The product's beginnings were relatively compact. Carriero built the engineering team of 15 that shipped the initial Dropbox for Business offering. His responsibilities subsequently expanded across the core Dropbox experience on multiple platforms. Eventually, he managed approximately 170 engineers across four groups. A product that began with a focused team had become part of a much broader management job.

A famous employer cannot do all your recruiting

When Carriero joined Segment as vice president of engineering in 2015, he encountered a less convenient starting point. Facebook and Dropbox were recognizable to almost any engineer. Segment sold customer data infrastructure to businesses. An applicant could be a capable programmer and have little reason to know its name.

The early recruiting process reflected that constraint. The team looked for contributors to open-source libraries and tried to persuade them to join. Carriero described the first year as uneven. Building an engineering reputation became a company priority, with technical writing, events and experienced hiring managers helping to give candidates a clearer picture of the work.

Over roughly his first two years, the engineering team grew from about 10 to about 55. Those numbers describe separate moments in one organization, rather than a comparison with the much larger group he had managed at Dropbox. They also explain why hiring was central to the job: every new colleague could materially change how the team worked.

When the product and the business disagree

At Segment, his responsibilities widened to include product, design and security alongside engineering. The chief product development officer title put several disciplines under one leader. It also brought him closer to questions that a clean technical architecture cannot answer by itself. Which problem deserves a team? What outcome would justify the investment? When should a promising project be stopped?

He wrote about one uncomfortable version of that problem in 2019. Six months into a new Segment product, the team was asking whether it had spent too long with too little to show. An established company had resources a startup lacked, but that made the stopping rule harder to see. A new product could keep consuming attention without the unmistakable deadline of an empty bank account.

This is a recurring difficulty in his work: extra capability can obscure a decision. More engineers do not decide what matters. More funding does not establish demand. More customer data does not tell a salesperson what to do next. Someone has to translate the available information into a choice that another person can act on.

His own accounts of management leave room for correction. A former Dropbox manager told him he wore his heart on his sleeve and needed to be calmer. He took the advice seriously. In discussing leadership, he has emphasized listening to concerns even when a day has already been crowded with other problems. Authority, in this telling, includes the capacity to hear an inconvenient answer.

“What good is collecting customer data if you don't trust it?”

Tido Carriero, Synapse 2018

Twilio acquired Segment in 2020 for $3.2 billion. Carriero had helped lead its product development through a period that ended with that transaction. The number belongs to the company's acquisition, rather than to his personal finances. The experience he carried forward was more specific: how a developer-facing product acquires the organizational machinery that enterprise customers require.

The analytics playground needed directions

In 2021, Carriero and his co-founders started Koala. Netto Farah and Matthew Shwery were fellow Segment veterans. Their starting observation concerned sellers: valuable information about prospective customers existed, but the people trying to help those customers often lacked the relevant context at the relevant moment.

Koala connected activity that could otherwise remain scattered across systems. Carriero's technical writing went into the details of identity and data models. A visit to a website, for example, meant more when connected to an existing profile and other activity. The engineering challenge lay in making those connections dependable enough to support somebody else's judgment.

But the product also taught its makers something about appetite. They initially imagined an analytics playground in which sales representatives would explore opportunities. Users wanted guidance about where to look. An expanse of possible discoveries could feel rather like being handed a map with every road highlighted.

Koala 2.0 reflected that lesson. Its Plays, Magic Inbox and Coach features aimed to connect signals with suggested next steps, research and personalized outreach. The emphasis moved toward reducing the work required to decide what deserved attention. The distinction matters to anyone designing a tool: making information available and making it usable are different tasks, with different demands on the person at the keyboard.

A company can end while the team continues

Koala's story did not end with the product becoming a permanent fixture in sales software. The company announced that its capabilities would sunset on September 30, 2025. It arranged refunds for unused subscription months and worked with Common Room on a transition for customers. The announcement acknowledged that Koala was no longer the team's answer to the changing demands of the market.

That ending deserves its place in the story. Customers had workflows to move and subscriptions to unwind. An acquisition could offer a destination for the builders while still leaving a product to close. Treating those as separate obligations gives a clearer view of what happened than simply adding another corporate name to a career timeline.

The destination was Cursor. Koala's team had used the coding tool while developing its own AI features. Now Carriero, Farah and Shwery would help it serve enterprise customers, bringing experience in security, collaboration and the practical demands of large organizations. Familiar problems had returned in a different product category.

Their stated priorities included permissions, audit trails, reliability and a connection between product adoption and enterprise sales. These are the details that let an enthusiastic individual user become part of an organization-wide deployment. A developer can choose a favorite editor in an afternoon. A company usually brings more people to the decision, some carrying spreadsheets.

The next bottleneck

By 2026, Carriero was speaking publicly as Cursor's vice president of engineering. His work again crossed the boundary between making a tool faster and making a larger system work. In May, he presented on agent workflows across software development. His responsibilities at Cursor also spanned product and design. Work with Atlassian connected planning context with coding agents, bringing the question of useful context into the software team's own workflow.

His public career advice offers a useful clue to how he evaluates these moves. He looks for meaningful impact, leaders he respects and the possibility of an outcome that changes a career or life. He has also advised Oso, drawing on his experience of how much effort authorization systems can absorb inside a growing software company.

In August 2026, he announced another transition as Cursor became part of SpaceX. He is now building at SpaceXAI. The setting has changed considerably since a small team was turning Dropbox into a business product. The question running through his work remains recognizably human: what does the person trying to use this actually need next?