This thesis presents a PoC for a Machine Learning based Integrated Business Forecasting tool in Integrated Business Planning (mature S&OP) environment for long-term and mid-term transportation volume forecasting for outbound supply flows as a case study.   To achieve this goal, the thesis provides a literature review which ensures the research gap presented by the research questions and scope. Later, the thesis introduces Sales and Operations Planning process, Machine Learning methods used in the study and the case study scope and the corresponding data minning scope. Following a mixed research strategy, a cause effect diagonis is presented and relevant business factors influencing the transportation volumes per lane are identified based on As-Is business process understanding achieved from interviews and internal and external documentations; which is further used to develop a conditional LSTM based machine learning time series forecasting model for transportation volume forecasting for five transportation lanes as PoC. Furthermore, a benchmark evaluation of the developed ML time series forecasting model with two other forecasting models (XGBoost and Extra Trees) is performed for accuracy and robustness performance metrics for long-term transportation volume forecast and also the performance of the developed ML forecasting model for the mid-term forecast is reported.  From the benchmark evaluation the proposed conditional LSTM model is proven to be a better balanced models in terms of maintaining acceptable level of forecasting accuracy and robustness. The Extra Trees model is the most accurate model with least robustness in forecasting across the five transportation lanes due to it’s inability to learn conditionally for each of the five transportation lanes.