24/7 conversational assistant for the Learning Planet Institute. RAG architecture powered by site content.
The system is based on a Retrieval Augmented Generation (RAG) architecture: site content is crawled, indexed in a vector database, then dynamically mobilized to generate contextual responses.
1. Crawling: Automated extraction of web content (pages, news, educational resources).
4 Generation: LLM synthesizes a response framed by the institutional system prompt.
The system prompt defines the tone, values and limits of the assistant: factual answers based solely on indexed content, redirection to a human for complex requests, respect for the institute's editorial line. This guarantees answers in line with the corporate identity.
24/7 availability for international visitors, reduced burden on the support team, improved accessibility to information. The chatbot handles recurring questions and directs specific requests to the appropriate contacts.
