Feb 2025 - Jun 2025
ScheduleAI is a personal final project developed for the course "Generative AI: Text and Image Synthesis Principles and Practice" It aims to transform natural language event descriptions into structured scheduling tasks through LLM-based reasoning.
-
Natural Language Event Parsing Users can input descriptions like "exam next Wednesday afternoon," and the system will automatically schedule related tasks such as review sessions.
-
Semantic Understanding via Google Gemini Utilizes LLM capabilities to extract time-related information, perform semantic normalization, and handle vague or implicit time references.
-
Chain of Thought Reasoning Employs multi-step logical reasoning to improve the accuracy of complex date/time inference and conflict resolution.
-
Prompt Engineering Designs carefully crafted prompts to guide the LLM in understanding intent and extracting precise scheduling details.
-
Extensible API and Front-End Integration Potential Built with modular architecture to support future front-end integration or API exposure.
Gantt Chart Support via Plotly for visualizing multi-task timelines and dependencies.
This project was completed as an individual final project for the course:
"Generative AI: Text and Image Synthesis Principles and Practice"