Hadi Partovi on Leadership, Hiring and Teaching Computer Science

Hadi Partovi tells Harry Stebbings that his route into startups began with his brother’s company LinkExchange and then his own departure from Microsoft to co-found Tellme Networks. After years as a technology operator, angel investor and adviser, he founded the education nonprofit Code.org.

Partovi says he repeatedly saw companies struggle to hire talent and concluded that schools were not preparing enough people for a technology-shaped future. Code.org’s mission, as he describes it, is to change school curricula so more students can learn computer science. He leads the organization and has a direct interest in that mission.

He grew up in Tehran during revolution and war, remembering nights in the basement during bombing and fears for his mother under morality policing. After the family emigrated to the United States, they were poor and shared one bed. Financial security became an early motivation.

Partovi describes his later aim as expanding opportunity for others. He says his family had been wealthy before the Iranian Revolution and that losing those resources gave his ambition an additional sense of reclaiming what had been taken. His story is his personal account of those events.

On the sale of Tellme to Microsoft, Partovi credits Emil Michael’s negotiating work. His own contribution, he says, was opening a conversation with Steve Ballmer, whom he had known since his early years at Microsoft. Tellme had another potential buyer and needed Microsoft at the table.

During a run, Partovi told Ballmer that Tellme and the talent Microsoft had once competed to recruit were available for a discussion. Ballmer asked to be connected. Partovi recalls colleagues later joking that he should no longer be allowed to go jogging with Ballmer because the encounter bypassed layers of process.

Partovi believes Ballmer deserves more credit for Microsoft’s earnings growth during his leadership, even though its stock performance at the time was less celebrated. He contrasts that assessment with Satya Nadella’s later work. This is Partovi’s interpretation of those periods, not a full financial comparison.

In product reviews, Partovi says both Ballmer and Bill Gates knew enough detail to challenge the supposed experts. That could create fear, but it also pushed teams to prepare thoroughly. He recalls Ballmer remembering an old product metric and asking a team to explain what had changed.

Leaders must balance delegation with involvement. Partovi admits he struggles with that balance but prefers to be close to the details, without doing everyone else’s job. In his view, executive scrutiny should establish a high bar and accountability.

He describes Gates as more focused on technology and product, and Ballmer as more focused on numbers and sales. Ballmer’s animated energy, Partovi says, was also motivating in person. Partovi recalls preparing well for those reviews and advancing rapidly at Microsoft.

Partovi first met Mark Zuckerberg in 2005 when Facebook had nine people and was receiving a term sheet from Accel. He recalls being struck by Zuckerberg’s seriousness and sweeping vision for identity, connection and a platform, rather than merely a site for college students.

Partovi became an adviser and says his arrangement valued the company at $10 million, compared with the $100 million valuation associated with the term sheet. He helped recruit across several functions, including introducing engineer Ari Steinberg. His investment and advisory relationship is relevant to his account of Facebook.

The first five or ten hires have an outsized effect because they shape not only the work being done but the people hired later. Partovi looks for exceptional ability alongside passion and commitment to the mission, across every role, not just engineering.

Recruiting those people is a founder’s central sales challenge. Partovi compares it with fundraising: a venture capitalist commits money, but a prospective employee may commit several years of their life and can choose only one employer at a time.

Discussing layoffs and higher interest rates in late 2022, Partovi argues that easy capital had allowed some weak spending decisions to persist. If layoffs are necessary, he favors making a sufficiently deep cut once rather than repeated rounds that undermine morale and leave remaining employees wondering who is next.

He questions whether startups should always raise as much cash as possible amid inflation and higher rates. Stebbings counters that securing funds can reduce uncertainty. Partovi acknowledges that concern but thinks very early startups may face greater risks from execution, co-founder relationships and finding product-market fit than from macroeconomic changes.

Partovi calls himself a workaholic and associates high performance with working both hard and smart. He says he does not demand his own hours from everyone else. As he has aged, he has paid more attention to physical and mental fitness because he believes that balance improves his performance.

He says he has not personally experienced burnout and feels energized by work he loves, contrasting it with externally imposed work. That is his individual experience and view; the interview does not establish that meaningful work protects everyone from burnout.

On speed, Partovi emphasizes quick decisions when sufficient information is available and errors can be reversed. He recalls eliminating meetings whose sole purpose was preparing for another meeting at Microsoft, favoring action over unnecessary process.

He also recognizes limits to moving fast. When an organization is larger and decisions have harder-to-reverse consequences for society, he says, leaders must slow down and consider those effects. He describes finding that balance as something he is still learning.

At Code.org, Partovi wrote down his decision-making framework so colleagues would not have to guess what he wanted. His leadership team practices hypothetical scenarios using the framework; they may reach different answers, but should share an understanding of the goals and tradeoffs.

The framework includes long-term impact, the cost of achieving that impact and the roles of partners. He gives a high-school computer science example: he aims for roughly $10–$20 per student, so a much more expensive intervention might produce less impact per dollar. This is an organizational target, not a verified universal cost.

One of Partovi’s largest personal decisions was to step away from maximizing wealth and measure his work more by societal impact. Around age 40, he chose that direction before he knew exactly which cause he would pursue; Code.org followed as a nonprofit.

He acknowledges that ego and comparison did not disappear. He still notices when peers become wealthier or more visible, but tries to distinguish his own measure of self-worth from what others reward. Stebbings points out that making such a choice is easier after financial success, and Partovi agrees.

Partovi says that if he still worried about rent or housing, he might remain focused on earning more. Even after changing direction, he feared that his social-impact work might fail. He credits Code.org’s reach with giving him greater confidence that the work mattered, though the reach figures are his organization’s claims.

On education, Partovi argues that schools should question what they teach as well as how they teach it. A child entering school at the time of the interview would begin working in the late 2030s, and he expects translation and AI tools to change some of the skills people use.

Stebbings worries that redesigning education around an uncertain technological future is a gamble. Partovi replies that keeping curricula unchanged is also a bet. He does not propose abandoning reading, mathematics or languages; he wants educators to reconsider priorities from first principles.

He argues for computer science, statistics, financial literacy and civics, and for learning that is more student-driven, creative, collaborative and project-based. His preference for computer science is connected to the mission of the nonprofit he leads, and his claims about future relevance are forecasts.

Asked how AI-generated code changes the case for teaching programming, Partovi says tools have long raised the level of abstraction. He expects generative AI to make instructions more natural, but argues that people will still need enough understanding to inspect whether an output does what they intended. This was his prediction in late 2022.

On retraining adults, he expects future work to become more digital and creative, making digital skills, problem-solving and creativity important. He argues automation will change some jobs and may free people to fill shortages such as teaching and nursing. He presents this as a view of possible transitions, not a measured employment outcome.

In the quick-fire round, Partovi names Sapiens as a favorite book. He is not planning to return to a for-profit operating role but would not rule it out. Looking back, he wishes he had taken larger bets on himself despite the fear of failure.

He recalls declining an opportunity to join Facebook before Sheryl Sandberg was hired because he was building a startup with his twin brother and had just raised money. He describes the decision as a choice of family loyalty over a potentially better financial outcome; the interview’s wording about the exact role is inconsistent.

Partovi praises Code.org vice chair Jeff Wilke for making time to help and for teaching through questions instead of simply giving answers. A good board member, he says, helps leaders develop their own solutions.

He says he would be interested in working at Axon, a public company whose board he serves on, because he sees potential for both growth and social impact in its law-enforcement technology. His projection of its growth potential and assessment of its products reflect his role and are not independent evidence of outcomes.

Partovi predicts that the next five to ten years will be exceptionally good for humanity. He argues that negative news receives disproportionate attention and points to long-term improvements in prosperity. These broad claims and his optimistic forecast are his perspective at the time, not verified forecasts.

For fatherhood, he emphasizes setting an example and making children feel loved and accepted. Looking five years ahead, he hopes the number of US states requiring computer science for high-school graduation will rise from five, as he reported at the time, to 25. The figure and target are his account and aspiration. Stebbings thanks him and ends the interview.