README.md

metadata

title: EEE_AI
emoji: 💬
colorFrom: yellow
colorTo: purple
sdk: gradio
sdk_version: 6.5.1
app_file: app.py
pinned: false
hf_oauth: true
hf_oauth_scopes:
  - inference-api

🎓 EEE_AI: Specialized Pedagogical Agent

Powered by Llama-3.1-8B-Instant & Groq LPU™

EEE_AI is an AI-driven tutoring system designed to assist students with the fundamentals of Electrical and Electronics Engineering (EEE). This project serves as a bridge between high-performance LLM engineering and domain-specific educational technology.

---

🚀 The Technical "Core"

As an AIML student, I built this space to explore the limits of low-latency inference and specialized guardrails.

• Inference Engine: Optimized via Groq LPU™, achieving speeds of 500+ tokens per second, making the tutoring experience feel instantaneous.
• Model: llama-3.1-8b-instant — chosen for its high-reasoning capabilities within a compact parameter count.
• System Architecture: A hybrid deployment using Hugging Face Spaces for the Gradio frontend and Groq Cloud for backend compute.

🎯 Key Functionalities

• Domain Focus: Provides structured explanations on Circuit Theory, Semiconductor Devices, and Power Systems.
• Interactive Tutoring: The model is prompted to act as a Socratic tutor—asking follow-up questions to test user understanding rather than just giving answers.
• Safety Guardrails: Includes a custom instruction layer that prevents the model from deviating into non-engineering topics, ensuring it remains a dedicated study tool.

🛠️ Tech Stack

• Language: Python
• Interface: Gradio
• LLM Framework: Groq API / Meta Llama 3.1
• Deployment: Hugging Face (Syncing with GitHub)

📖 Sample Interactions

User: "Explain KVL in simple terms." EEE_AI: [Provides explanation] + "Would you like a practice circuit problem to test this law?"

User: "What's the best movie to watch tonight?" EEE_AI: "I am focused on your engineering success! Let's get back to EEE—perhaps we can discuss how Signal Processing is used in movie audio instead?"

---

👷 About the Developer

Kushagra Gaur | Curious from Core This project was developed to master the integration of high-speed inference APIs and the implementation of domain-specific constraints in LLMs.