🎓 Coursebot: University Agentic Platform

Coursebot is an AI-native platform designed for university staff to architect, deploy, and monitor specialized lecture bots. Built to bridge the gap between static lecture materials and active student inquiry, Coursebot allows educators to transform course syllabi into interactive, "grounded" knowledge agents.

---

🏛️ About the Project

Coursebot is a reference implementation for an Agentic Educational Ecosystem. It demonstrates how modular AI agents can be managed within a university framework to provide real-time, pedagogically sound guidance to students.

The platform provides a centralized hub for:

• Staff-Led Bot Architecture: Non-technical creation of lecture-specific agents.
• Knowledge Grounding: Ensuring AI responses are strictly anchored to verified course DNA.
• Interaction Observability: Real-time analytics into student lexical patterns and bot efficacy.

---

🚀 Core Capabilities

This application demonstrates high-level agentic workflows, including:

• DNA Sequencing (Mistral AI): Automated extraction of course metadata (ECTS, outcomes, faculty) from raw academic catalogs.
• Vectorized Knowledge Vaults (Weaviate): Hybrid search indexing of lecture materials to prevent hallucination.
• Granular Student Enrollment: Protocol-driven access management ensuring students only interact with authorized coursebots.
• Interaction Analytics: A visual dashboard for staff featuring:
  • Inquiry Word Clouds: Identifying trending student topics.
  • Frequency Mapping: Tracking the 15 most common pedagogical keywords.
  • Conversation Inspection: Full markdown-rendered conversational logs for quality assurance.

---

🛠️ Technical Architecture

Coursebot is built on a modular stack designed for scalability and secure data handling:

• Brain (LLM): Mistral AI (via API)
• Memory (Vector DB): Weaviate (Hybrid Search & GraphQL)
• Backend: PHP 8.x (Secure Proxy Pattern)
• Frontend: Vanilla JavaScript, Chart.js, Marked.js
• Protocol: Custom JSON-based Agent Interface

---

📦 Getting Started

Prerequisites

• PHP 8.1+
• A Weaviate Cluster (Cloud or Local)
• Mistral AI API Credentials

Installation

1. Clone the Repository:

   git clone [https://github.com/YourUsername/coursebot.git](https://github.com/YourUsername/coursebot.git)
   cd coursebot

2. Initialize Configuration:
   • Copy api/config.php.example to api/config.php.
   • Insert your Weaviate and Mistral credentials.
3. Deploy to Server:
   • Point your web server to the root directory.
   • Ensure the UserRegistry and CourseRegistry classes are initialized in your Weaviate cluster.

---

⚖️ License

Distributed under the MIT License. See LICENSE for more information.

Project maintained by Philip Lau.