The project was developed within Politecnico di Milano with the aim of experimenting with the application of artificial intelligence technologies to support administrative processes. These processes can often be time-consuming and require a considerable amount of expert resources. The project focuses on the creation of a conversational chatbot dedicated to student services, designed to provide precise answers on topics related to students’ academic careers. The chatbot relies on a structured knowledge base, which is regularly updated by the student services staff. The ultimate goal is to reduce the time staff members spend providing accurate responses, thereby improving the overall experience of students using the service.
Collection and structuring of sources
Development of a knowledge base to support the chatbot, built from existing documentation mainly available on the website and in various PDF documents. Design and implementation of a software system capable of processing the knowledge base stored in a cloud folder and converting it into a database accessible by the chatbot.
Deliverables:
- Organized knowledge base.
- Software system for interaction.
- Vector database for the chatbot.
Chatbot development and deployment
Software implementation of the chatbot and its related pre-processing tools, with the release of a demo accessible via web and API for experimentation.
Deliverables:
- Code for the pre-processing pipeline and chatbot functions.
- Working demo accessible online.
Testing and training
Qualitative testing of the chatbot by student services staff, with support for troubleshooting any issues. Creation of a file for the structured collection of feedback. Organization of workshops aimed at transferring knowledge both to administrative staff and to ASICT technical staff responsible for maintaining the software in operation.
Deliverables:
- Feedback and observations collected during the testing phase, organized in a file.
- Technical documentation of the software systems’ functioning.
- Training workshops on chatbot development and maintenance.
The project is currently in the experimentation phase. Expected outcomes include more efficient handling of recurring requests by student services, faster response times for users, and the acquisition of internal technical and organizational skills to maintain and further develop the tool over time.
Co-Director
AIRIC
Nicola
Gatti
AI Research Scientist
AIRIC
Piergiuseppe
Pezzoli
AI Research Scientist
AIRIC
Davide
Fabroni
AI Research Engineer
AIRIC
Andrea
Chizzola
AI Research Scientist
AIRIC
Tomaso
Castellani
Director of ICT Services Division
ASICT | Politecnico di Milano
Fabrizio
Pedranzini
Head of Digital Transition Strategy and IAM Units
ASICT | Politecnico di Milano
Roberto
Gaffurri
Head of Students Affairs Unit
ASED | Politecnico di Milano
Gianluca
Attolini
AI Engineer
ASICT | Politecnico di Milano
Davide
Mozzi
