AI-powered video captioning that understands short clips (30s–2min) and generates four distinct styles — formal, sarcastic, humorous-tech, and humorous-non-tech — plus emotional learning insights for parents and teachers.
This application does not provide medical or psychological diagnosis. It offers AI-assisted early observations to help parents and teachers provide better support.
Detect 13 emotional categories from video analysis
Identify learning barriers before they escalate
Tailored recommendations for parents and teachers
Parents or teachers upload a short 30-second to 2-minute classroom or home video.
Fireworks AI multimodal model analyzes facial expressions, body language, speech, and interactions.
Fireworks LLM rewrites the scene summary in formal, sarcastic, humorous-tech, and humorous-non-tech tones.
“EduSense AI helped me understand my daughter's classroom anxiety weeks before her teacher noticed. The recommendations were practical and kind.”
Sarah M.
Parent
“The teacher strategy suggestions are spot-on. I adjusted my classroom approach and saw real improvement in a shy student's participation.”
James T.
Kindergarten Teacher
“A thoughtful early support tool — not a diagnosis, but a valuable conversation starter between parents, teachers, and support staff.”
Dr. Lisa K.
School Counselor
No. EduSense AI provides AI-assisted early observations to help parents and teachers provide better support. It does not replace professional evaluation.
We support MP4, MOV, and AVI files up to 2 minutes in length and 100MB in size.
We use Fireworks AI's multimodal video models (Qwen3 Omni) to analyze facial expressions, body language, classroom interactions, and speech patterns.
Videos are processed for analysis and reports are stored locally in your browser. No data is sent to third parties beyond the Fireworks AI API for processing.
Per hackathon Track 2 requirements, we generate formal, sarcastic, humorous-tech, and humorous-non-tech captions for each analyzed video.