How Peak Ji Built Manus into a Global AI Sensation
Viral debut and rapid user growth
When Yichao Peak Ji appeared in a simple launch video for Manus in March, few expected the reaction it sparked. Filmed in a corner of Butterfly Effect’s Beijing office and directed by cofounder Zhang Tao, the clip showcased Ji speaking in fluent English and explaining the AI agent. Within days the product, then an invite-only preview, spread from China to the wider world and drew a waiting list of roughly 2 million people.
What Manus does differently
At first glance Manus looks like a typical chatbot: users type questions into a chat window. But Manus goes further by executing tasks on behalf of the user. It breaks a request into steps and runs those steps on a cloud virtual machine equipped with a browser and other tools, allowing it to browse websites, fill forms, and complete multi-step tasks such as finding an apartment that fits specified criteria and budget.
Ji’s role and the team
Yichao Peak Ji is the technical core of Butterfly Effect and serves as chief scientist. Now based in Singapore, he leads product and infrastructure development as the company pushes to expand globally. His cofounders Zhang Tao and Xiao Hong bring complementary strengths in product, storytelling, and organization, forming a team that mixes deep technical ability with go-to-market savvy.
A history of building products
Ji has over a decade of experience building consumer and enterprise products that combine technical depth with usability. He first gained attention as a teenager creating iOS apps, including the Mammoth browser, which became a top third-party iPhone browser in China and won the Macworld Asia Grand Prize. At 20 he appeared on the cover of Forbes and made its 30 Under 30 list. He later launched Peak Labs and built Magi, a search tool that used a custom language model to surface answers from across the web. Magi attracted millions of users initially and found a second life as an enterprise product before being sold in 2022.
Open source roots and product instincts
Raised in a household of academics and IT professionals, Ji learned programming early and was influenced by open-source culture, Bill Gates, and Linux. Teenage projects and early successes taught him product design and risk management, including an experience with a legal notice over a Monopoly-branded app that sharpened his instincts rather than deterring him.
Global infrastructure and ambitions
Manus is built on a global stack: US-based infrastructure and technologies such as Claude Sonnet, Microsoft Azure, and open-source tools like Browser Use. With new funding led by Benchmark, the team relocated key operations to Singapore and is actively targeting international consumers. The product is a good example of modern AI entrepreneurship, where talent, tools, and infrastructure cross borders rapidly.
A founder looking ahead
Despite rapid growth and international interest, Ji frames Manus as a potential capstone product. He says he hopes Manus is the last product he needs to build, because it would be able to handle any future wild ideas he might have. For now, he and his team are focused on scaling the agent and turning early virality into lasting global adoption.