Heilbronn University is researching advanced technical tools and didactic approaches designed to foster collaboration and interaction. The primary goal is to enhance and facilitate interdisciplinary, inter-faculty, and inter-university cooperation. Through the InduKo research project, we aim to strengthen this collaboration using innovative digital tools and modern, digitally supported strategies that boost motivation and foster innovation through partnership.
Academic institutions are full of untapped collaboration potential. Researchers, faculty, students, alumni, and industry partners often have complementary interests but struggle to find the right connections. Lucii helps bridge this gap by enabling targeted networking, whether for research projects, study groups, mentoring, or industry collaboration. Lucii is a specialized matching solution designed to connect individuals in academic environments based on shared interests, expertise, and collaboration potential. Built on the Matrix messaging protocol and powered by AI on Edge, Lucii enables privacy-preserving matchmaking and initial anonymous contact without requiring a centralized database or external data processing. All profile evaluations and interest matching take place locally on the user’s device, ensuring that no sensitive profile information is stored on central servers.
Lucii showcases how AI on Edge and the Matrix protocol can bring together the right people in a privacy-preserving way. By combining decentralized messaging with local AI-based matchmaking, Lucii demonstrates how meaningful academic connections can be established without the need for central processing and storage of highly personal profile information.
InduKo Project was funded by Stiftung Innovation in der Hochschullehre (August 2021 - July 2024).
Prerequisites
To set up Lucii, ensure you have:
Flutter SDK (for app development)
Dart SDK (included with Flutter)
Android Studio or VS Code (for testing on an emulator or physical device)
Matrix Dendrite Server (for decentralized messaging)
Gemini Nano API (for on-device AI)
Installation
- Clone the Repository:
git clone cd Lucii
- Install Dependencies:
flutter pub get
- Run the App:
flutter run
Lucii Features in Detail
- AI-Powered Profile Matching
Users provide profile information in natural language.
Gemini Nano extracts keywords to categorize user interests.
Profiles are matched locally without sending data to external servers.
- Post Creation & Discovery
Users can create posts related to research interests.
Posts are tagged using on-device AI, ensuring relevance and personalization.
Users can browse and engage with posts based on personalized recommendations.
- Anonymous Messaging
Users can reply to posts anonymously via Matrix.
Initial messages remain private until users opt to reveal identities.
Conversations occur securely within the decentralized Matrix network.
- On-Device AI Processing
Gemini Nano processes user input locally, ensuring full privacy.
AI refines user profiles, adapting to engagement patterns.
No centralized data collection occurs.
- This project is licensed under Apache 2.0. Copyright 2024 , Hochschule Heilbronn.
This project was developed as part of the InduKo Project, funded by Stiftung Innovation in der Hochschullehre. We also acknowledge the support from students, faculty, and contributors who have been part of this collaborative effort.
For more information about the InduKo research project, visit the official website.