Forecasting cash flow data

Project description

The main challenge of this project was to design a reliable and understandable cash flow forecasting solution using accounting data from Moneybird. The central research question was how financial data could be extracted, processed, and visualized in a way that supports management decision making, while also allowing different forecasting methods to be tested and compared. A key part of the challenge was dealing with irregular transaction patterns, limited historical data, and the need for automation. The project focused on translating raw accounting data into meaningful insights through forecasting models and a dashboard that balances analytical depth with clarity and usability for non-technical stakeholders.

Context

This project was conducted in the financial domain and focused on cash flow forecasting for InnoVactions, a company that currently relies on manual spreadsheet-based forecasting. Financial data such as sales invoices, purchases, and bank transactions are stored in Moneybird, but extracting and using this data for forecasting was time-consuming and difficult to maintain.

The project took place as a Proof of Concept, with the goal of exploring whether a more automated and visual solution could improve insight into cash inflows, cash outflows, and future bank balances. The domain required careful handling of financial data, attention to data quality, and transparency in forecasting logic, as the results would be used to support management decisions.

Results

The main outcome of the project is a validated Proof of Concept that demonstrates how financial data can be transformed into an automated and transparent cash flow forecasting solution. A Power BI dashboard was developed that combines historical figures with forecasted values, providing a clear monthly overview of cash flows and future trends.

In addition to the dashboard itself, the project delivered an automated data flow solution. Financial data is extracted through Python scripts and stored in a structured format in a shared folder, which acts as a lightweight data layer for Power BI. By using parameters inside Power BI, the data source can be easily adjusted without changing the underlying logic, making the solution flexible and maintainable.