-
-
AI-powered mentor evaluation platform homepage
-
Mentor performance dashboard with AI analytics
-
Lecture analysis with parameter-wise insights.
Inspiration
Traditional mentor evaluation systems are often manual, inconsistent, and unable to measure accessibility and inclusiveness in teaching. We wanted to create an AI-powered solution that not only evaluates teaching quality but also ensures that education becomes more accessible for differently-abled learners such as deaf, visually impaired, and speech/hearing challenged students. Our inspiration came from the idea that great teaching should be understandable and inclusive for everyone.
What it does
MentorMind-AI is an AI-powered mentor evaluation platform that analyzes recorded teaching sessions using multimodal AI. The system evaluates communication clarity, engagement, technical depth, confidence, interaction quality, and accessibility. It generates AI-based mentor scores, detailed feedback reports, accessibility insights, and comparative dashboards to help institutions improve teaching quality and inclusiveness.
How we built it
We built the frontend using React, Tailwind CSS, and modern dashboard components for an intuitive user experience. The backend and AI workflow integrate Gemini API for intelligent analysis and report generation, Whisper for speech-to-text transcription, and OpenCV/MediaPipe for visual and engagement analysis. The system processes uploaded teaching videos, extracts transcripts and audio patterns, evaluates mentor performance across multiple parameters, and generates detailed AI-driven evaluation reports.
Challenges we ran into
One of the biggest challenges was designing fair evaluation metrics because teaching styles vary greatly between mentors. Another challenge was analyzing accessibility-related parameters such as speech clarity, pacing, and inclusive communication for differently-abled learners. Integrating video, audio, and text analysis together into a single workflow while maintaining performance and accuracy was also challenging within limited development time.
Accomplishments that we're proud of
We are proud of building an inclusive AI-based mentor evaluation system that goes beyond traditional scoring methods. Our platform not only evaluates teaching effectiveness but also focuses on accessibility and learner inclusiveness. We successfully combined multimodal AI analysis, accessibility-focused scoring, and automated feedback generation into a single working prototype with a clean and modern user interface.
What we learned
Through this project, we learned how multimodal AI systems can be used in real-world educational applications. We gained hands-on experience in AI-powered video analysis, speech processing, accessibility-focused product design, API integration, and rapid prototyping. Most importantly, we learned how AI can help create more inclusive learning experiences for all students.
What's next for MentorMind-AI
In the future, we plan to add real-time mentor evaluation during live sessions, sign language recognition support, emotion-aware engagement analysis, multilingual accessibility features, and institution-wide analytics dashboards. We also aim to build personalized mentor improvement recommendations powered by AI to help educators continuously enhance their teaching effectiveness and accessibility.
Log in or sign up for Devpost to join the conversation.