Chapter 4 — The Startup Path: From 0 to 1
The Idea Is Born
In junior year Li Mingxuan learns about AI’s potential in medical diagnosis. Moved by stories of long waits and misdiagnoses, he proposes an AI-assisted diagnostic platform. The team debates feasibility, privacy, and regulatory hurdles.
Market Research and Resources
They use UBC’s library and partnerships, attend industry forums, interview clinicians, and study data privacy regulations (PIPEDA, HIPAA). They find strong government support and usable technical approaches like federated learning.
UBC entrepreneurship programs, Hatch incubator, e@UBC accelerator, and Sauder resources help them get started.
Team Building and Product Prototype
Team roles: - Wang Xiaojie — AI engineer - Li Mingxuan — product and architecture - Zhang Haoran — business development - medical advisors and graduate data scientists
Tech stack: React front-end, Python/Django back-end, TensorFlow/PyTorch models, distributed storage, and cloud deployment. Core features include imaging analysis, symptom triage, drug interaction checks, and decision support for clinicians.
They iterate with agile development, secure anonymized hospital data for training, and use techniques to improve accuracy and performance. Initial accuracy meets a high bar in pilot tests.
Funding, Legal, and Growth
Funding sources: UBC grants, seed funds from competitions, government programs (IRAP, Mitacs), and angel investors.
Legal work includes IP strategy (patents, trademarks), company registration, and regulatory compliance with Health Canada and privacy laws.
They achieve pilot success, secure angel funding, and face challenges: regulatory timelines, long sales cycles, cash flow, and hiring competition.
Culture and Next Steps
Li cultivates mission-driven culture, hires from campus, uses equity incentives, and plans A-round financing and market expansion while keeping focus on social impact.