[Unitaryhack] Build a Documentation Q&A Chatbot for Quantify

Summary: Create a chatbot that can answer questions about Quantify’s documentation. This tool should help users and contributors quickly find relevant information, improving their onboarding and development experience.

Why this matters: As Quantify grows, high-quality, accessible documentation is critical. A chatbot interface can lower the barrier to entry for new users and reduce time spent searching for answers in the docs. This increases community engagement and developer velocity — exactly what we need to scale our ecosystem. Completing this bounty helps you showcase AI integration skills and provides a valuable tool for all current and future users of Quantify.

In scope:

• Set up a chatbot that can:
  • Be queried via a local script or a simple web interface
  • Run locally or be hosted on our website
  • Answer questions using our documentation (#26 (closed)) as a knowledge base
• Use an open-source LLM or embedding-based retrieval method (e.g., LlamaIndex, LangChain, Haystack, OpenAI embeddings if free tier is sufficient)
• Support questions like:
  • “What are the different abstraction layers of quantify?”
  • “What is a schedule?”
  • "What different available pulses can I apply to a qubit?"
• Provide setup instructions (README) and optionally a short demo video or screenshots

Out of scope:

• Training a custom LLM
• Deep integration with the GitLab UI
• Real-time chat services (e.g., Discord/Slack bots)
• Deployment to production or enterprise-grade hosting

Deliverables:

Merge Request containing:

• Working prototype of the chatbot
• Instructions on how to run it
• Code must be documented and easy to follow

Reference the issue in your MR and write Closes #[issue-number] in your MR description

How to Win: We’ll prioritize completeness, clarity, and ease of use

Bonus Points:

• Clean UX/UI
• Fast response times
• Configurable indexing pipeline (e.g., easy to reindex when docs change)