ClassCatch AI

• 

  ClassCatch AI

• 

  ClassCatch AI-Architecture

Inspiration

Young students often miss out on important learning—especially children aged 7–10 who have recently immigrated and are still learning German. When they are absent or cannot fully understand the lesson, they fall behind quickly. Teachers want to support these students, but they don’t have time to create personalized summaries every day, and parents often don’t know how to help at home.

We created ClassCatch AI to make learning accessible for every child, no matter their background or language level. A fun, friendly avatar explains the day’s lesson in simple terms, helping kids catch up and helping families stay connected to the classroom.

What it does

ClassCatch AI automatically turns a teacher’s daily lesson materials into a child-friendly, animated learning summary.

Here’s what the system does:

Teachers upload notes, PDFs, images, or recordings to Google Drive

The AI summarizes the lesson using age-appropriate, simple language

A 3D avatar (e.g., “Mousey”) explains the lesson through animation and voice

Students can listen or watch their summaries anytime through the ClassCatch app

Parents get easy access to the content to help their children at home

Multilingual support ensures understanding even for non-native speakers

A gamified interface makes students excited to review their lessons

ClassCatch AI transforms traditional classroom material into a fun daily learning experience.

How I built it

ClassCatch AI is powered by a fully automated orchestration pipeline built using n8n, Google Drive, LLM services, Supabase, and cloud storage.

Ingestion Layer

Google Drive Trigger detects new teacher uploads

PDF extractor pulls lesson content

AI model (Gemini) generates summaries and scripts

Supabase stores all text data

Production Layer

Parent clicks “Generate” → n8n Webhook activates

System fetches script from database

Text-to-speech service creates kid-friendly narration

Animation API produces the avatar lesson

Audio and video are uploaded to cloud storage (S3)

Student & Parent Layer

Study portal for parents

ClassCatch app for students with audio/video streaming, XP progress, and gamification

This pipeline ensures the entire process—from upload to animated lesson—is smooth, automated, and scalable.

Challenges I ran into

Making AI simplifications truly understandable for 7- to 10-year-olds

Ensuring the summaries stayed accurate while still using child-friendly language

Building a clean workflow in n8n that handles multiple steps without errors

Syncing audio and animation outputs from separate APIs

Managing multilingual content in a consistent and culturally neutral way

Creating an interface simple enough for teachers, yet engaging enough for kids

Each challenge helped improve the product and pushed the architecture to be more robust.

Accomplishments that I'm proud of

Building a fully automated pipeline from teacher upload → AI summary → animated lesson

Creating adorable, engaging avatar lessons that kids genuinely enjoy

Designing a tool that supports immigrant families and improves accessibility

Implementing a reliable end-to-end architecture using modern tools (n8n, Supabase, S3, Gemini)

Making something that reduces teacher workload while improving student comprehension

Turning complex lessons into something fun, friendly, and easy for kids

What I learned

How to design LLM prompts specifically for young learners

How to build automated multi-step workflows using n8n

Best practices for voice generation, animation APIs, and media streaming

How to create educational tools that are both engaging and inclusive

The importance of focusing on user experience for teachers, parents, and children

How multimodal AI (text + audio + animation) can dramatically enhance learning

What's next for ClassCatch AI

Adding AI-generated interactive quizzes based on daily lessons

Allowing students to ask the avatar questions in real time

Introducing new animated characters

Expanding multilingual output to support more families

Developing a full mobile app for easier at-home access

Partnering with schools to conduct real-world testing and refine the system

ClassCatch AI is only the beginning—our goal is to make daily learning accessible, fun, and personalized for every child.

Built With

• ai
• ar
• canva
• gemini
• lovable
• mixamo
• n8n
• slidesgo