AI Time Series Pattern Recognition

I want to automate the analysis of my time-series data so that recurring patterns surface instantly and with reliable accuracy. This project is pure data analysis—no chatbots or customer-service flows—so everything you build should focus on ingesting streams of dated records and uncovering the structures that repeat, shift, or fade over time. I will supply the raw data in CSV format along with a short data dictionary. Your task is to: • clean and pre-process the series, • develop or fine-tune an AI model (Python with pandas, scikit-learn, TensorFlow, or PyTorch are all fine) that spots statistically significant patterns, and • generate a clear report and reusable code that reproduces the results on future datasets. Acceptance criteria • Model identifies known historical patterns with a precision/recall above 0.85. • Scripts run from the command line with one config file and no manual tweaks. • Output includes a markdown summary, plots, and the saved model weights or pipeline for deployment. If that workflow sounds straightforward, let’s get started—I’m ready to review your approach and timelines.