BeSmart Campus needed a way to deliver individualized instruction at scale after demand for tutoring surged, but static online video could not answer questions or adapt explanations to each student's pace and understanding.
Reported impact
BeSmart launched its first operational platform in three months, produced measurable student gains in school tests, and created a scalable tutoring model that can re-explain concepts as many times as each learner needs.
BeSmart Campus built an interactive learning platform on Google Cloud so educational video content can be paired with real-time AI tutoring instead of passive playback. The company uses Gemini models to analyze and organize teacher-created video lessons, then relies on multimodal understanding to let students pause a lesson at any moment, ask a question about the exact point on screen, and receive an explanation that follows BeSmart's own pedagogical method. Vertex AI and Gemini supply the reasoning layer, while BigQuery supports the data foundation behind the platform's evolving content and results. The system also generates practice exercises on the fly and uses the same course context to explain mistakes, rephrase concepts repeatedly, and guide students through alternative examples. This architecture turns the platform into a personalized tutor that can keep pace with each learner while staying inside a controlled instructional environment.
Teacher-created videos are analyzed and organized with Gemini, students watch those lessons in the BeSmart platform, pause them to ask context-specific questions, and receive grounded responses or generated exercises that stay tied to the exact lesson segment and teaching method.
Dutch EdTech company BeSmart Campus leverages Google Cloud and Gemini models to offer every student a personalized AI tutor, directly integrated into educational content designed with teachers.
Thanks to its multimodal understanding, Gemini immediately identifies the relevant section and provides a response consistent with the pedagogical method designed by BeSmart Campus.
The AI never gets impatient and can rephrase ten, twenty, or thirty times if necessary.
The first operational version of the BeSmart Learning Platform was designed in only three months
two classes allowed to use the platform for two weeks preceding exams achieved averages of 6.4 and 6.9 out of 10, while two other classes of the same level, prepared without the platform, showed an average of 5.0
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
The Organization JSON-LD lists numberOfEmployees value 13, and sameAs 'http://www.besmartcampus.com'.
Related
Google Cloud Sample Modified · Oct 9, 2026 Mindvalley Mindvalley needed to scale personalized AI learning companions for a large active-user base while keeping latency, uptime, and compute costs under control across customer-facing conversational workloads. 1.3 million annual agent requests BigQuery · Cloud Run · Cloud SQL Google Cloud Sample Archived · Jun 6, 2026 Neutopia However, the platform’s previous cloud architecture required ongoing DevOps management, placing pressure on a lean engineering team. More significantly, the surrounding ecosystem was optimised for technical efficiency… Gemini · Gemini Enterprise Agent Platform · Imagen Google Cloud Sample Archived · Apr 29, 2026 University of California, Riverside University of California, Riverside is a global engine for social mobility, helping students from all backgrounds move into high-earning careers. Ranked in the top 1.3% of universities worldwide, UCR empowers over… Reduces administrative document tasks from multiple hours to minutes with Gemini BigQuery · Gemini · Google Cloud Storage 