A high-performance, full-stack ChatGPT clone designed as a specialized AI Assistant for institutional policies. It leverages modern web technologies and advanced AI orchestration to provide users with an intuitive chat interface capable of retrieving and reasoning over complex university or organizational documentation.
Purpose
Serves as an "Institutional Policy Specialist." Unlike general-purpose LLMs that may hallucinate or lack specific internal data, this clone uses Retrieval-Augmented Generation (RAG) to ensure responses are grounded in actual policy documents. It is designed to help students, faculty, and staff navigate institutional rules with precision, conciseness, and accuracy.
Key Features
AI Agent with Memory
Uses LangGraph to manage conversation state, allowing for multi-turn dialogues where the AI remembers previous context and maintains coherent conversations.
Semantic Search (RAG)
Integrates ChromaDB to perform vector searches across policy documents, providing relevant excerpts to the LLM for accurate, evidence-grounded responses.
Policy-Specific Tools
Specialized tools including search_policies, list_all_policies, get_policy_by_name, and summarize_policy for navigating institutional documentation.
Persistent Chat History
Messages and threads are saved to a PostgreSQL database, allowing users to return to previous conversations and maintain context across sessions.
Feedback System
Built-in mechanism for users to rate AI responses, facilitating continuous improvement and monitoring via LangSmith for quality assurance.
Modern UX
A responsive, dark-mode-first UI that mimics the ChatGPT experience, including message streaming and sidebar navigation for seamless interaction.