Question-Smith

A flexible interactive notebook aimed at helping users build a tool that takes technical questions and returns beginner‑friendly explanations using intelligent language models.

---

Repository Structure

├── LICENSE
├── Project 4.ipynb
└── README.md

LICENSE

This is the standard Apache License 2.0, granting rights to use, modify, and distribute the project under specified conditions. It provides the legal framework for contributions, derivative works, and redistribution. ([uithub.com][1])

Project 4.ipynb

A Jupyter notebook designed as a guided exercise for users learning to build an AI-powered tool. It walks through:

• Setting up environment variables and verifying the OpenAI API key.
• Defining model constants like gpt-4o-mini and llama3.2.
• Crafting system prompts to roleplay a programming tutor that explains technical concepts in beginner‑friendly language.
• Implementing a mechanism to stream responses from an OpenAI-like model via yield from.
• Supporting both OpenAI and Ollama backends for inference flexibility.

This notebook structures the technical scaffolding while also demonstrating best practices for clean prompt engineering and API integration. ([uithub.com][1])

---

Overview & Purpose

question-smith is intended to be an instructional and foundational tool for anyone looking to learn how to:

• Interface with large language models (LLMs) like OpenAI’s GPT series and open-source alternatives like LLaMA.
• Engineer system and user prompts to create a task-specific assistant (a programming tutor, in this case).
• Implement streaming responses for responsive and interactive output.
• Facilitate beginner-friendly explanations, breaking down complex code into easy-to-understand steps.

---

Target Audience

Ideal for:

• Educators and learners in prompt engineering or AI-integrated programming education.
• Developers building tools that convert technical queries into accessible explanations.
• Users exploring hybrid architectures that blend proprietary models (OpenAI) and open-source frameworks (Ollama).

---

Key Features

• API Key Validation: Sets up and validates environment variables to guarantee smooth integration and user feedback. ([uithub.com][1])
• Model Flexibility: Supports both GPT-based proprietary models and local LLaMA models via Ollama.
• Prompt Templates: Includes system prompts defining tutor behavior and user prompts for simplicity and clarity.
• Streamed Responses: Implements a streaming approach with yield from for dynamic, responsive output.
• Beginner-Focused: Prioritizes clear, jargon-free explanations using Markdown formatting.

---

How to Get Started

1. Clone the repository and open Project 4.ipynb in a Jupyter environment.

2. Set your OpenAI API key using a .env file or environment variable (e.g., OPENAI_API_KEY).

3. Run through the notebook step by step:

   • Validate the API key and environment setup.
   • Adjust or experiment with system prompts to fine-tune the teaching style.
   • Try both the OpenAI and Ollama backends to compare responses.

4. Explore and extend:

   • Change the prompt to tutor in different languages, disciplines, or difficulty levels.
   • Swap out models or add more logic (e.g., caching, error handling, advanced prompt structures).
   • Integrate with other interfaces or UIs (e.g., web apps, chat interfaces).

---

Why It Matters

This project offers a practical, hands-on introduction to constructing AI-powered tutoring tools:

• Bridges coding and pedagogy: Demonstrates how to structure prompts that teach as well as compute.
• Models flexibility: Showcases how to switch between commercial and open-source LLMs.
• User-first clarity: Emphasizes compassionate, clear communication—essential for teaching technical topics effectively.

---

License

Distributed under the Apache License 2.0, allowing for broad use and contribution under defined terms. ([uithub.com][1])

---

Contributing

Interested contributors can:

• Improve the tutorial prompt structure for different learning styles.
• Add support for additional models or backends (e.g., local CPU-only setups, new open-source LLMs).
• Enhance usability by integrating visualization tools or interactive front-end components.
• Expand explanation templates to cover more domains beyond code (e.g., math, data science, algorithms).

---

Summary Table

Component	Purpose
LICENSE	Legal usage and distribution terms
Project 4.ipynb	Main educational notebook and experimentation playground
README.md	High-level orientation and usage guidelines

---

Feel Free to Contribute by improving this Question-Smith