Squirrel Ai’s Joleen Liang outlines AI education strategy at Harvard

Squirrel Ai co-founder Joleen Liang returned to Harvard University for a third student session, where she discussed AI-driven personalized learning, the company’s growth, and its U.S. expansion strategy. The appearance highlighted how the Chinese edtech company is positioning its adaptive learning platform for global markets and schools. Why it matters: - Squirrel Ai is using a high-profile Harvard platform to pitch its model for AI-driven personalized education to future business, education and policy leaders. - The company is framing its approach as a response to a core education problem: scaling personalized learning without losing precision. - Liang’s remarks also point to how Chinese edtech companies are trying to navigate regulatory shifts at home while expanding abroad. What happened: - Dr. Joleen Liang, co-founder of Squirrel Ai, delivered an invited session at Harvard University for the third time. - The event combined a one-on-one interview format with an interactive Q&A. - Dozens of master’s and doctoral students attended from Harvard Business School, Harvard Graduate School of Education and Harvard Kennedy School. - The discussion covered AI in education, Squirrel Ai’s commercial growth and the company’s global strategy. The details: - Liang said Squirrel Ai began in 2014, before AI tutoring became a mainstream topic. - She said the company focused early on using AI to address the limits of standardized education and improve educational equity. - Squirrel Ai built adoption by emphasizing measurable learning gains and personalized experiences. - The company’s intelligent tutoring platform has served more than 50 million users. - Liang said China’s 2021 tutoring-industry regulatory changes forced the company to adapt while staying focused on its mission of empowering education through AI. - Squirrel Ai’s systems are now integrated into more than 60,000 public schools. - Liang said the company shifted from pursuing scale to deepening educational and ecosystem value. - Liang described Squirrel Ai’s Multimodal Adaptive Education Large Model as different from general-purpose large language models. - She said the model uses micro-granularity knowledge graph parsing, dynamic personalized learning path planning, knowledge root-cause tracing and capability framework building. - Liang said Squirrel Ai’s Large Action Model can diagnose learning gaps and adjust a student’s learning path in real time. - Liang said AI should free teachers from rote instruction and grading so they can focus on creativity, critical thinking and emotional mentorship. - As head of Squirrel Ai’s North American expansion, Liang said the company is using a franchise model in the U.S. instead of a pure software-as-a-service rollout. - She said the model is designed to combine AI efficiency with local, high-touch service. Between the lines: - The Harvard appearance gives Squirrel Ai a chance to position itself as more than an edtech vendor; the company is presenting its model as a system for human-AI collaboration in classrooms. - Liang’s emphasis on long-termism suggests Squirrel Ai wants to distinguish itself from short-cycle edtech ventures that depend on rapid user growth. - The franchise approach in the U.S. signals that Squirrel Ai expects local execution to matter as much as the technology itself. What’s next: - Squirrel Ai is likely to keep pushing its human-AI collaboration model as it expands in North America. - Liang’s comments suggest the company will continue using schools, localized operators and adaptive learning tools to support global rollout. - The company is also likely to keep presenting its platform as a China-developed alternative to generic AI tools in education.