MapleMate is a personalised AI assistant, built by MLIS Robotics Club of Maple Leaf International School, Dhaka. It is based on Google's Gemma3 model from Ollama, designed to assist with various tasks such as answering questions, providing information, and engaging in conversations. It is being gradually tailored to give students and teachers a helpful tool for various academic and administrative tasks, and assist with their learning and teaching processes and be insync with the school.
P.S. Win goes to Gemma3, it runs on a single GPU and it is still pretty fast!
- Internet Access: Enabling real-time information retrieval from the web.
- Containerization: Packaging the application in a Docker container for easy deployment.
- Integration: Connecting with school systems for real-time information access.
- Multilingual Support: Expanding language capabilities to cater to a diverse user base.
- Vision: Incorporating computer vision for image analysis.
- Voice Interaction: Enabling voice commands and responses for hands-free operation.
- Andrej Karpathy for the inspiration and guidance in building AI assistants.
- Ollama for providing a proper framework to run the Gemma3 model.
- Google's Gemini Gemma3 for the powerful and lightweight language model that powers MapleMate.
- Tashinur Rahman Tashin for the website UI and logo design for MapleMate.