Last weekend, I had the incredible privilege of serving as a mentor at CSE Summer School 2025, hosted by the University of Technology – Ho Chi Minh City (HCMUT). The theme this year was “Agentic AI” — a concept that’s both cutting-edge and deeply inspiring. I was fortunate to guide a team of bright, passionate high school students as they explored, built, and brought to life their first intelligent systems.
I was assigned to guide Team 7, a group of energetic and talented high school students, as they explored how AI agents can solve real-world problems through intelligent reasoning and interaction.
From Curiosity to Capability
From the very beginning, I was impressed by how eager and curious the students were. Many of them had only basic programming knowledge, but that didn’t stop them from diving headfirst into AI concepts like reasoning, planning, and autonomy.
Over the course of two days, I introduced them to tools such as OpenAI’s GPT APIs, LangChain, and especially Pydantic — a powerful Python framework that allowed them to structure their AI agents cleanly and efficiently. Pydantic proved to be an excellent bridge between abstract AI logic and concrete implementation, giving the students clarity as they modeled memory, dialogue states, and user profiles.
Mentorship Beyond Code
As a mentor, I didn’t just want to teach syntax or architecture. I wanted to help the students believe that they could build real-world AI systems. I shared stories from my own projects, explained how agentic AI is being used in industry, and encouraged them to ask bold questions.
Watching them go from tentative to confident in just 48 hours reminded me why mentorship is so important. It’s about planting seeds — not just of knowledge, but of mindset and possibility.
A Real Project: VaxNOVA – An Agent for Preventive Healthcare
By the end of the first day, our team had converged on a clear, meaningful direction: building a health-focused AI agent named VaxNOVA. The goal was to assist users in recognizing symptoms, receiving personalized care advice, and most notably, getting vaccine recommendations based on individual factors like age, gender, and condition history.
While it was just a prototype, the system demonstrated thoughtful flow design, responsible data use, and a public-good orientation. VaxNOVA wasn’t just a chatbot — it reflected a new way of thinking: “How can we use AI to empower people to take control of their health?”
The team used Pydantic to define schema models for user data, symptom inputs, vaccine info, and agent memory — making the system’s logic interpretable, testable, and extensible.
Reflections and an Open Invitation
If there’s one thing this experience reinforced, it’s that mentorship matters — not just in transferring technical skills, but in helping young minds see themselves as builders, not just consumers, of technology.
Using tools like Pydantic reminded all of us that even the most powerful AI agents start from thoughtful design, structure, and clarity. AI is not about chasing hype — it’s about making things work in the real world, for real people.
I truly believe this kind of exposure should happen more often, especially at the high school level. These students are the future of AI — and after this weekend, I’m hopeful.
As the program came to a close, what stayed with me the most weren’t just the lines of code or the working prototypes — it was the energy, the collaboration, and the shared sense of discovery we experienced together. These two days have become a cherished memory, a reminder of how powerful learning can be when driven by curiosity and guided with care.
To my incredible Team 7 — I hope you always carry this momentum forward. Keep building, keep questioning, and never be afraid to reach beyond your limits. The spark I saw in each of you deserves to grow.
A heartfelt thank you to the organizing committee, the research team, and everyone behind CSE Summer School 2025. Thank you for giving us — and me — this beautiful opportunity to grow, connect, and be inspired. I’ll carry this experience with me for a long time.