I want a Python-based workflow that starts with gathering historical stock prices and ends with a working deep-learning model that can generate forward-looking signals. The data should come from Prowess CMIE, NSE/BSE feeds, and—where gaps exist—Yahoo Finance. Once collected, the prices must be cleaned, merged, and stored in a reproducible format (CSV or a lightweight database is fine). After the data pipeline is in place, I’d like you to design and train a predictive model—preferably a deep neural network—using well-known libraries such as pandas, NumPy, scikit-learn, and either TensorFlow or PyTorch. The emphasis is on producing an end-to-end, repeatable process that I can run locally to refresh the dataset and retrain the model. Deliverables • Python scripts or notebooks that pull, clean, and save stock price data from the stated sources • Model-building code with clear comments, tuned hyperparameters, and a concise validation report (MSE/RMSE or comparable metric) • A short README explaining setup, required API keys, and exact run commands If you have prior experience working with Indian exchanges or deep learning for financial time series, please highlight it when you respond.