On June 19, 2025, a team of undergraduate freshmen from URA (Undergraduate Research in AI), successfully delivered a Question-Answering (QA) system to Vexere, one of Vietnam’s leading online travel platforms. What makes this accomplishment especially meaningful is that the entire system was built from scratch by first-year students and is now officially being deployed in Vexere’s production environment.
This was not a capstone, a coursework assignment, or a sandbox experiment. It was a real-world project, with real users, real business constraints, and production-level standards. Our task was to design an AI-powered assistant that could support users in key booking processes, including:
- Changing or canceling booked trips (one-way and round-trip)
- Searching for journeys and retrieving trip information
- Initiating payments through Vexere’s backend systems
From day one, this project demanded more than technical skills. We had to collaborate closely, think like product engineers, adapt quickly to changing APIs, and ensure our solutions aligned with business workflows – all under tight deadlines and real production pressure. Delivering this project meant navigating uncertainty, learning continuously, solving integration issues, and responding rapidly to feedback. Despite the complexity, we maintained a strong sense of ownership and professionalism.
Throughout the journey, our team worked across multiple domains – from AI modeling to API design, backend service development to system debugging. Our progress was made possible by collaboration, shared effort, and a deep belief in learning by doing. At URA, we believe in empowering students early with real responsibilities and real impact. More than just sharpening our technical skills, this experience has built confidence, resilience, and a mindset of growth. We’re excited for what comes next.
The first-year student team with Assoc. Prof. Quan Thanh Tho and Nguyen Song Thien Long