Cracking the Clock: Unlocking Smart Predictions for Effortless Task Completion

Accurate effort estimation at the task level is essential for effective project planning, resource allocation, and meeting delivery timelines in software development. While traditional approaches focus on project-level estimation, limited research addresses predicting the duration of individual tasks. This paper presents a novel hybrid machine learning approach for task effort estimation by integrating Google’s Gemini API with Facebook’s Prophet time series model. The proposed system uses task-specific features—such as name, priority, complexity, labels, and start date—to predict completion times, combining semantic understanding with temporal modeling. Experimental results using real-world Jira datasets demonstrate the feasibility and superior accuracy of this method compared to traditional estimation techniques. This work bridges the gap between high-level project estimation and fine-grained task-level forecasting, offering a data-driven solution to enhance agile development planning.