David Zhu’s company is trying to give sales teams a longer memory. Some of the people building it have been part of his life since childhood. Before there was a product called Reevo, before the funding announcement or the argument about artificial intelligence, there were classmates who kept turning up in one another’s lives. Silicon Valley likes to discuss networks. A school friend who becomes a colleague, then a co-founder, is a rather less abstract proposition.
That history gives Zhu’s story an unusual symmetry. His business asks what an organization loses when knowledge sits in separate places. His founding team brings together people with years of shared context. The connection is an interpretation, but a useful one: knowing what happened before can change how quickly people understand what is happening now.
The friends came first
Clement Fang has known Zhu since elementary school. Curtis Tan met him in sixth grade; they later attended the University of California, Berkeley, and worked at DoorDash. Cindy Hao’s connection came through OpenStore, where she and Zhu led the Growth team. The four founded Reevo in 2024.
A founder introduction can resemble the beginning of a negotiation. These introductions had already happened. That does not settle a disagreement about a product or make a deadline easier to meet. It does give the disagreement a history longer than the deadline. A person can be wrong on Tuesday without becoming a stranger on Wednesday.
Zhu was born in China and moved to the United States in 1997. He credits his father’s entrepreneurial example with an early interest in building things. Growing up in the Bay Area, he questioned standardized schooling and says he was tutoring calculus by eighth grade. There is something pleasingly practical about that response: while questioning the educational system, he also found someone who needed help with the homework.
An engineer notices what gets lost
Zhu’s career took him into engineering leadership. At DoorDash, his remit included New Verticals, Platform, Money, and Enterprise. He built and led teams of more than 200 engineers across three sites. Those are responsibilities involving both software and the people who must keep it working as an organization expands.
His later explanation of Reevo begins with a familiar engineering problem: crucial knowledge can live inside a particular person’s head. When that person leaves, colleagues discover how much of the system was never truly shared. Documentation helps only if the useful details reach it, remain current, and can be found when someone needs them.
Zhu sees a similar vulnerability in sales. An experienced representative knows which objections matter and how a customer reached a decision. A database may preserve the account while failing to preserve the understanding. His ambition is to make software carry more of that context. The challenge is as much about continuity as convenience.
A company milestone, rather than a measure of Zhu’s personal wealth.
The money brought room to build. It also brought an obligation to make the broad proposal work in the ordinary hours of somebody else’s working day. A software thesis eventually has to survive a user who simply wants to finish a task before lunch.
Draw the work before drawing the product
Reevo’s preparation included mapping the jobs of different revenue roles and the connections between them. Marketing, sales development, account executives, revenue operations, and customer success all participate in a customer’s journey. Looking at their work as a graph made the handoffs visible before the team chose its product focus.
The method suits an engineer. A complaint about one screen may describe a problem that begins two departments away. Following the work through the organization changes the unit of analysis. Instead of asking only whether a feature is good, the builder asks what happens to the information after the person using that feature is finished.
That wider view could invite an indiscriminate roadmap. Reevo’s approach also distinguished discovering needs, building for them, and selling to a segment. Zhu described a focus on growing companies rather than immediately pursuing the largest enterprises, where trust requires a track record. A startup can accelerate a release. It has fewer options for accelerating the passage of time.
conversation↓Shared
context↓A useful
next action
The trouble with a very well-informed fragment
Zhu’s writing returns to a specific objection: putting AI inside separate applications can leave each application ignorant of the others. A call recording may contain an important hesitation. An email exchange may explain it. A pipeline record may carry neither. Giving every fragment a fluent voice does not automatically reunite the conversation.
His proposed architecture puts prospecting, outreach, and deal information on a common foundation. The aim is for a useful answer to draw on the customer’s wider history, then help someone act. That is his product argument, with all the responsibility that such an argument places on execution.
In March 2026, he made the setup burden itself a subject of his writing. Sellers, in his view, should be able to begin selling without first assembling a collection of applications. The point is easy to appreciate even for someone perfectly happy with their existing tools. Buying software creates work of its own. Every connection needs to keep carrying the right information.
There is an almost domestic quality to the complaint. You bought the appliances. You have become the electrician. Zhu wants the assembled system to be the product, including the less glamorous work between its parts. Whether that arrangement earns a customer’s loyalty depends on how reliably it helps when a real deal gets complicated.
A specialist with permission to wander
His hiring philosophy follows the work across similar boundaries. Zhu has described wanting specialists who also investigate the broader customer problem. Engineers, designers, and marketers retain their crafts, while contributing ideas, researching, and interacting with customers. The job description establishes a starting point for attention.
The appeal is straightforward. Someone who understands a problem closely may notice a useful change outside their formal area. The difficulty is equally clear: shared curiosity needs enough discipline to become a decision. Otherwise, a room full of interested people can produce a very elaborate conversation and nothing a customer can use.
“We attack problems, not people.”
David Zhu
Zhu describes learning from mistakes as part of the culture he wants. Taken seriously, that principle puts a demand on the leader as well as the employee. It asks the person with authority to examine an unsuccessful decision without making the next experiment feel personally expensive. Optimism becomes credible through what happens after something goes wrong.
The pitch meets a customer
A concrete example comes from Propel founder Tony Dong, whom Zhu pitched on Reevo. Dong had an engineering background of his own, including leadership at Rippling, and was building an AI product. Running sales alongside that work had left him moving between spreadsheets, messages, and other records to reconstruct a deal’s status.
Propel reported importing its spreadsheet data and starting to use the platform within a day. Dong used a Slack bot to add opportunities and ask about deals. The attraction was partly architectural, but the useful result was mundane: fewer small administrative steps standing between a conversation and an updated record.
That example brings Zhu’s proposition down to scale. Nobody has to contemplate the future of an entire software category to understand why an opportunity should be easy to record. The customer is entitled to judge the grand design by the small errand. In business software, the small errands have an unfortunate habit of multiplying.
A bigger platform, a slower kind of persuasion
On July 8, 2026, Reevo announced its acquisition of Ciro, a prospecting agent that finds, qualifies, enriches, and drafts outreach to sales contacts. The company said the technology would strengthen prospecting inside its existing platform. For Zhu, the acquisition extends the argument that finding the next customer should happen alongside managing the relationship.
The same announcement put Reevo above 90 employees. The organization building the shared platform was becoming larger itself. That gives the founding ideas another practical test: preserving context within a team becomes more demanding as more people join it. An approach that feels natural among old friends must become intelligible to new colleagues.
Zhu has also publicly reconsidered how a new category gains support. In a LinkedIn reflection, he acknowledged underestimating the role of other people’s voices. He argued that an idea spreads through people willing to carry it into conversations of their own. The implication is humbling for any founder with a persuasive product presentation. You cannot attend every room in which your company needs to make sense.
That observation returns the story to relationships. Software can make information available; people still decide what deserves their attention and trust. Zhu is working on the machinery that helps a company remember. The test ahead is whether the remembered details make the next conversation better. A long memory earns its keep through what somebody does with it.