←back to thread

128 points Soumame | 2 comments | | HN request time: 0s | source

Hello! I’m So Tokumaru ( https://tokumaru.work ), high school student from Japan working on ways to leverage large language models (LLMs) for K-12 programming education. Recently, I built a project called "TutoriaLLM," a self-hosted application designed to support educators and students through a multi-modal LLM that offers guided tutorials.

TutoriaLLM integrates voice (voice-to-voice) and text-based AI interactions into Blockly, a block editor widely known for its use in Scratch. It enables server-side code execution and allows students to test their code directly in Minecraft (Bedrock or Education Edition) without any setup.

I’m actively developing this project and aim to bring it to a level ready for use in programming education by next year. I’d love to hear any feedback!

Documentation: https://tutoriallm.com

Demo: https://demo.tutoriallm.com

Repository: https://github.com/TutoriaLLM/TutoriaLLM

Show context
amrrs ◴[] No.42075354[source]
This is impressive, the fact the dev is high school student is mind blowing! Curious what LLM is used and what kind of role LLM plays?
replies(1): >>42075369 #
1. amrrs ◴[] No.42075369[source]
Looks like it uses OpenAI - https://github.com/TutoriaLLM/TutoriaLLM/blob/bac952b288eded...

Would be interesting to see if we can replace it with openai-compatible local llm end point to run this locally

replies(1): >>42075754 #
2. Soumame ◴[] No.42075754[source]
Correct, I am using OpenAI through the software, includes, gpt-4o, 4o-audio-preview(for multi-modal response, includes audio and text), and whisper(speech to text for generating transcript for user).