Student Advisor

Organizing your studies can quickly become confusing for students—between emails, deadlines, and course materials in learning management systems such as Moodle. This is where the Student Advisor app comes in. It is an AI-based application designed to simplify everyday student life by connecting directly to external services such as email inboxes or platforms such as Moodle.

I developed this project as part of a university project in a team and received a grade of 1.0 for it. It is a project that honed my skills in backend development and AI integration.

The focus was exclusively on Python-based logic. A frontend was not developed.

I used Python as the programming language and relied heavily on Langchain and Langgraph to implement multi-agent networks. The architecture is based on these multi-agent networks, with the agents designed to communicate with each other to solve complex tasks.

For email processing, for example, a RAG (retrieval-augmented generation) approach was used, which utilizes a Chroma database.

Another key component was the Moodle agent: it was able to log in to Moodle (using the user's login details), search for information there itself, create files, and upload course materials.

Since the project was created before MCP became widespread, I had to resort to other solutions, such as **web scraping**, to communicate with Moodle.

Student Advisor was developed as a university project. It is therefore not deployed anywhere, is not actively used, and is currently not being further developed.

Through the implementation of the project, I was able to learn how to build complex multi-agent networks—i.e., the combination of LLM and program code—from scratch and how the Langchain and Langgraph libraries work.