Coursera and Its Founders: How Andrew Ng and Daphne Koller Reshaped the Global Online Education Industry

If you place Coursera in the history of online education, it was neither the earliest open-course platform nor the only MOOC company, but it is very likely the company that most successfully integrated university-branded content, career certificates, enterprise training, online degrees, platform distribution, and capital markets into one durable business system. It was launched by Andrew Ng and Daphne Koller in 2012 with the mission of giving “anyone, anywhere” access to world-class learning. Its real origin was the 2011 Stanford large-scale online course experiment, after which it quickly evolved from a “free-course platform” into a learning and skills infrastructure company.

In terms of scale, Coursera is no longer an “education idealist experiment.” It is a public company with M&A activity and a strong enterprise orientation. Official filings show that by the end of 2025, Coursera had about 197.3 million registered learners and more than 375 university and industry content partners; full-year 2025 revenue was $757.5 million, up 9% year over year, while net loss narrowed to $51 million. By the first quarter of 2026, registered learners had already reached 205 million. After combining with Udemy and deepening its AI strategy, the company described itself in July 2026 as a comprehensive skills platform reaching “more than 300 million learners.”

The relationship between the two founders is fundamentally asymmetrical but highly complementary. Andrew Ng is better understood as the narrative engine, AI education brand, and long-term strategic voice. He has remained chairman and continues to bind Coursera to the broader AI industry through DeepLearning.AI, AI Fund, LandingAI, his Amazon board seat, and his 2026 startup LearnVector. Daphne Koller, by contrast, is better understood as the academic authority, learning-vision architect, and early product-and-pedagogy driver. Her influence on course design, learning science, university partnerships, and the belief that technology could deepen learning was profound in Coursera’s formative years, but after 2016 she shifted to Calico, insitro, and Engageli, and is no longer an operating leader at Coursera.

In one sentence, Coursera’s real place in the world is this: it is both one of the most successful commercial survivors of the MOOC era and a representative case of how the ideal of “open education” gradually gave way to credentialization, enterprise monetization, subscriptions, and AI-driven learning. The outside world remembers it not only because it put university courses online, but because it helped push online learning away from passive content consumption and toward verifiable skills and career pathways.

Founders’ backgrounds, education, and personal trajectories

Public information on Andrew Ng is strongest on education and career history, while family details are relatively sparse. His official website confirms that he holds a bachelor’s degree from Carnegie Mellon, a master’s degree from MIT, and a Ph.D. in computer science from UC Berkeley. Stanford’s official page confirms his long career in AI research and teaching and his central role in both online education and artificial intelligence. As for family background, widely circulated public profiles say he was born in London in 1976, grew up in Hong Kong and Singapore, that his parents were from Hong Kong, and that he graduated from Singapore’s Raffles Institution. But his parents’ occupations, family class position, and more granular information about his childhood resources remain publicly limited / inconsistent / unconfirmed.

Andrew Ng’s educational path reflects the layering of mathematical ability, engineering training, academic research, and industrial deployment. From CMU to MIT to Berkeley, and later to Stanford AI leadership, this trajectory helps explain why he repeatedly occupies the position of asking how research can scale into real systems, rather than remaining inside purely academic work. The strongest influences on him appear not to be a single thinker but the intersection of machine learning, deep learning, internet distribution, and engineering at scale.

Andrew Ng’s earliest representative professional identity was not Coursera but Stanford AI research and Google Brain. Stanford’s official materials show that he founded and led Google Brain and also served as Baidu’s chief scientist. Coursera emerged from his 2011 online Stanford machine learning course, which attracted over 100,000 learners and made him ask whether education could scale like software. In other words, he entered education not because he began as an education-sector operator, but because he was an AI researcher who believed technical distribution could radically reshape the cost structure of learning.

Andrew Ng’s later entrepreneurial portfolio is strikingly coherent: Coursera addressed the large-scale distribution of high-quality knowledge; DeepLearning.AI addressed standardized AI skills training; LandingAI addressed industrial AI deployment; AI Fund addressed startup incubation; AI Aspire addressed enterprise AI strategy; and LearnVector is attempting to make learning itself AI-native. These ventures are not unrelated parallel assets. They are successive expansions around the same core theme: putting advanced AI capabilities into the hands of more people and organizations.

As an influence asset, Andrew Ng’s real strength is not only equity ownership or board seats, but the ability to connect courses, platforms, industry, capital, and public narrative. Coursera’s proxy materials show that as of March 2025 he beneficially owned roughly 7.2 million shares, about 4.5% of the company. They also show that Coursera maintained a platform and revenue-sharing agreement with DeepLearning.AI, an entity he controls, and paid related entities about $8.4 million in 2024. That means he is not a symbolic founder who drifted away. He remains deeply embedded in Coursera’s content ecosystem and strategic direction.

Daphne Koller’s public biography is more complete. Widely used public references record that she was born on August 27, 1968, in Jerusalem. Her Stanford and Engageli materials confirm that she earned her undergraduate and master’s degrees at the Hebrew University of Jerusalem, then completed a Ph.D. at Stanford and joined Stanford’s computer science faculty in 1995. A Coursera profile further states that she aspired from an early age to study at Hebrew University, began taking courses there while still in high school, had completed her master’s degree by age 18, fulfilled her Israeli military service, and then began doctoral work at Stanford.

One element of Daphne Koller’s family influence can be stated with relatively high confidence: her father, Professor Dov Koller, was an academic whose influence reached other scholars who later taught on Coursera. An early Coursera post explicitly referred to “Daphne Koller’s father, Professor Dov Koller.” Her mother’s profession, the family’s wealth level, and the exact structure of household resources remain publicly limited / not clearly confirmed. What is visible is that she grew up in a highly education-dense environment, which fits her exceptionally accelerated academic trajectory.

Daphne Koller’s academic standing is exceptionally strong. Her Stanford biography shows that she completed her Ph.D. at Stanford in 1993, did a postdoc at Berkeley, and joined Stanford in 1995. She later received a MacArthur Fellowship, entered the National Academy of Engineering, and built a career in probabilistic graphical models, machine learning, and computational biology. She did not move into online education because she had left scholarship behind. Rather, she had already spent years inside Stanford thinking about how technology could improve engagement and depth in learning. In an Engageli interview, she said that before Coursera she was already using technology to improve the educational experience for Stanford students.

Daphne Koller’s post-Coursera path is also highly continuous, though the sectors changed more dramatically. She left Coursera in 2016 to become chief computing officer at Alphabet’s Calico; founded insitro in 2018 to apply machine learning to drug discovery; and co-founded Engageli in 2020, bringing her back to the question of interaction quality in online learning. So she did not simply “leave education for biotech.” She has repeatedly worked on the same deeper problem: using computation to redesign highly complex systems.

In terms of founder archetypes, Andrew Ng is closer to an industrializing thought leader, resource connector, and founder-brand figure, while Daphne Koller is closer to a research founder, pedagogy shaper, and methodological architect. Coursera was able to combine public narrative, university legitimacy, and technical credibility in its 2012–2016 years precisely because those two modes of leadership were bound together early on.

Company architecture, business model, and capital relationships

Coursera did not begin with a company first and courses later. It began with Stanford’s 2011 large-scale online course experiment and only then became a company. Andrew Ng’s official site and Coursera’s tenth-anniversary history both confirm that his 2011 online machine learning course attracted more than 100,000 learners, and that in 2012 Ng and Koller launched Coursera with Stanford, Penn, Michigan, and Princeton as founding partners. That origin matters: Coursera began not with ad-driven traffic or user-generated content, but with elite university licensing and professor brands.

Coursera’s early product logic was “free large-scale courses + top-university content + centralized platform delivery and credential handling.” By the time of its 2021 S-1, the company clearly divided revenue into Consumer, Enterprise, and Degrees. Consumer included single-course certificates, Specializations, and catalog-wide subscriptions. Enterprise included Coursera for Business, for Campus, and for Government. Degrees covered full online degree programs. By the 2025 10-K, the reporting structure had shifted to two reportable segments, Consumer and Enterprise, while the platform still retained courses, Specializations, Professional Certificates, MasterTrack, and more than 50 degree offerings. That shift suggests that degrees remain important, but the company increasingly wants to be understood as a unified lifelong-learning and institutional skills platform.

Its business model has gone through several clear upgrades. It began by using free courses to build scale, then sold credentials; later it added Specializations, monthly subscriptions, and catalog subscriptions; in 2016 it launched enterprise products, moving the revenue center beyond consumers; by 2018 it used MasterTrack and degrees as stackable credentials; after 2019 it expanded Guided Projects, Campus, and Government offerings; and by 2025–2026 it was leaning into AI-powered coaching, role simulation, Course Builder, and Skills Tracks to improve retention, conversion, and institutional value. The 2025 filing explicitly says that Coursera Plus subscription growth was a major driver of Consumer-revenue growth.

If you separate “real assets” from “influence assets,” Coursera’s hard assets are not school campuses or a publishing house, but platform distribution, its registered learner base, the network of university and enterprise relationships, the content library, trusted brand equity, and long-term learning-behavior data. The 2025 10-K explicitly treats its roughly 197 million registered learners as a strategic asset because they drive Consumer revenue, attract content creators, generate Enterprise leads, improve SEO performance, and strengthen operating scale economics.

Capital-wise, Coursera has always looked like a classic Silicon Valley edtech company. It raised $16 million in Series A in 2012 from Kleiner Perkins and NEA; $43 million in Series B in 2013 from investors including GSV, the World Bank’s IFC, Laureate Education, Learn Capital, Yuri Milner, and follow-on participation from Kleiner Perkins and NEA; a Series C led by NEA in 2015; another $64 million in 2017; and a $103 million Series E in 2019 led by SEEK with participation from Future Fund and NEA. The significance of this capital history is that Coursera was never designed as a slowly expanding mission-first nonprofit. It was built from the beginning as a high-growth platform company.

In 2021, Coursera priced its IPO at $33 per share, sold roughly 15.73 million shares, and raised around $519 million, becoming a NYSE-listed company. More importantly, just before going public it converted into a Delaware Public Benefit Corporation and obtained B Corp certification. On the surface, that was a mission statement. In substance, it was also a governance design choice: the company was telling the market that it intended to preserve a public-learning narrative rather than operate solely as a short-term shareholder-maximizing vehicle. At the same time, its 10-K acknowledged that PBC status may require balancing investor returns against broader stakeholder and public-benefit interests.

The founder-company relationship has also remained economically active. Andrew Ng is both chairman and, through DeepLearning.AI, a course supplier to Coursera under a revenue-sharing arrangement. In July 2026, Coursera also invested $100 million into LearnVector, a new company founded by Ng, taking about a one-third fully diluted stake. That investment was approved by a special committee of independent directors. In other words, the company and its founder did not separate into neatly distinct domains after professional management took over; they have remained strategically and transactionally entwined.

In terms of long-term partnerships, Coursera has not been defined by a single media conglomerate. It has instead relied on four structural pillars: universities, technology companies, enterprise training budgets, and capital markets. Universities provide legitimacy, tech companies provide job-relevant certificates, enterprises and governments provide B2B revenue, and public markets provide financing and M&A instruments. Among its notable recent relationships are Google, IBM, AWS, Microsoft, and DeepLearning.AI. This is exactly why Coursera has steadily moved away from being just a “university course website” and toward becoming a broader skills-and-employability platform.

Turning points, controversies, current status, and timeline

The most important turning points in Coursera’s history were, first, the 2011–2012 move from Stanford experiment to incorporated company; second, the 2014 decision to bring in former Yale president Richard Levin as CEO, marking a shift from professor-led startup to institutional platform company; third, the 2016 launch of enterprise offerings, which made B2B and B2B2C central to the revenue thesis; fourth, the 2021 IPO together with PBC conversion; and fifth, the 2025–2026 phase of management transition, the Udemy combination, and the push toward AI-native learning. Each step reduced dependence on the original MOOC story.

The management evolution is particularly revealing. In 2014, when Rick Levin became CEO, Andrew Ng became chairman and chief evangelist while Daphne Koller became president. In 2016, Koller left for Calico. In 2017, Jeff Maggioncalda became CEO and notably intensified commercial scaling and IPO preparation. In 2025, Greg Hart took over as CEO, with the company emphasizing his background in technology-driven innovation and operating discipline. That same year Coursera launched an expense-reduction initiative expected to generate at least $30 million in annualized structural savings. The company has become increasingly similar to a public technology company centered on growth, margin discipline, integration efficiency, and AI strategy.

Coursera’s greatest success is not just scale, but the way it transformed “learning content” into a product system that can be stacked, verified, distributed, and purchased by institutions. It enabled university brands to enter global skills markets and let enterprise brands enter higher education and lifelong learning through professional certificates. Outside research suggests that online course and MOOC credentials are not meaningless in labor markets: one study found substantially higher employer preference for resumes containing MOOC credentials, and another experiment found that cost-free access to curated Coursera courses and certificates improved employment-related outcomes. Coursera did not eliminate the university; it changed how learning connects to work.

But its biggest controversies are concentrated in the same area. During the early MOOC wave, the most common criticism was low completion rates. Discussions in 2013 pointed to roughly 5%–7% completion in Coursera-style MOOCs. Andrew Ng and Daphne Koller responded in an EDUCAUSE piece by arguing that treating every registrant as if they had always intended to complete the course fundamentally misreads MOOCs, because many learners enroll only to browse or sample content. In other words, the deepest controversy around Coursera has never really been whether people use it, but whether this form of use should count as “real education.”

The second major controversy is the erosion of openness. Coursera originally became famous through the combination of “free, open, elite-university.” But in April 2025, the company’s official support center stated that the new model replaces the traditional audit experience with free preview access to only the first module of most courses. Longtime MOOC observers such as Class Central described this as another step away from the original open-education ethos. So today the most concentrated criticism of Coursera is not a singular scandal but that it no longer resembles the free global classroom it once symbolized, and increasingly resembles a paid skills platform. On this point, public opinion is genuinely split: supporters see sustainable monetization, while critics see the near-end of the MOOC ideal.

The third category of controversy is more typical of platform companies than defining moral crisis. Coursera’s 2025 10-K explicitly disclosed that the company had been party to a class action under the Video Privacy Protection Act and had faced arbitration demands and claims involving privacy, consumer protection, accessibility, advertising, marketing, and related areas. Separately, merger materials in 2026 show that the Coursera–Udemy transaction faced stockholder lawsuits and demand letters over disclosure issues. These look more like standard legal frictions for a public platform company than like fundamental scandals that define its history.

In terms of current status, Coursera is now in a phase defined by post-merger integration and AI restructuring. In May 2026, it completed its combination with Udemy and described the combined business as building the world’s most comprehensive skills platform. But in July 2026, Reuters reported that the company was cutting jobs after the merger in order to optimize cost structure and operating model. That means scale expansion has already happened. The next problems are platform integration, efficiency, and product differentiation.

Andrew Ng’s 2026 moves are especially important. Both official Coursera materials and Reuters confirm that Coursera invested $100 million into his new startup LearnVector and took roughly a one-third stake. The company framed the deal as a major plank of its AI strategy and as a bet that AI expands the learning market rather than shrinking it. The significance is large: Coursera is not treating AI merely as a feature layer inside the platform. It is trying to use the founder’s next-generation AI learning company to reshape its own future. The founder and the company have entered another tight strategic loop.

Daphne Koller’s current influence, by contrast, is now largely outside Coursera itself and concentrated in insitro and Engageli. Official biographies identify her as founder and CEO of insitro and as co-founder and board member of Engageli. Insitro announced a $400 million Series C in 2021, and Engageli announced a $33 million Series A in 2021. Today she is better understood as a research-driven repeat founder operating across AI, edtech, and life sciences than as a present-day power center inside Coursera.

A compressed timeline looks like this: Andrew Ng’s online machine learning course explodes in 2011; Coursera is founded in 2012; Rick Levin becomes CEO in 2014 while the founders remain influential but less operational; enterprise learning becomes a major pivot in 2016, the same year Daphne Koller exits; Jeff Maggioncalda takes over in 2017 and accelerates commercialization; the IPO, PBC conversion, and B Corp status land in 2021; Greg Hart becomes CEO and pushes cost discipline in 2025; Coursera combines with Udemy and places a major strategic bet on Andrew Ng’s LearnVector in 2026. At this point Coursera has moved from being a star startup of the MOOC era to being core skills infrastructure for the AI era.