VibeCheck is a platform I designed and built to create a safe, low-stakes space for students to reflect on assignments while giving instructors meaningful, aggregated insights about student learning experiences.
The Problem
Instructors design assignments with specific learning goals, but rarely have a structured way to understand whether students experienced the assignment as intended. Traditional course evaluations are too infrequent and too broad. VibeCheck closes that gap by creating a lightweight feedback loop tied to individual assignments.
How It Works
- Instructor Intentions — As they finalize an assignment, instructors archive their intended learning goals. Takes less than 3 minutes.
- Student Reflection — Students record brief reflections as they submit the assignment. Takes less than 2 minutes. All responses are anonymized and aggregated.
- Learning Insights Report — Student data is aggregated and analyzed against instructor expectations, producing an actionable report on alignment, performance themes, and areas to address.
- Institutional Trend Analysis — Raw data is stored for cross-course, longitudinal, and institutional analysis, tracking macro changes in learning and student experiences over time.
Design Philosophy
VibeCheck is grounded in Situated Learning and the idea that trust is a prerequisite for collective learning. The tool scaffolds trust through:
- Mutual engagement — Instructors and students reflect on the same elements (task, goals, strategies, AI use)
- Joint enterprise — A low-stakes way to ensure course goals and intended experiences are shared
- Shared repertoire — A common tool and mutual reflection process producing actionable insights
By protecting anonymity, VibeCheck encourages honest communication and gives instructors timely insights into how their expectations align with student experiences — opening space for dialogue and iteration.
Technical Stack
- Backend: Python/Flask with Google Cloud Run
- Frontend: React
- Data: Google Sheets (privacy-first — student identities and responses are never stored together)
- AI: LLM-powered insight generation for Learning Insights Reports