Data has rapidly become an important driver for innovation. They are often presented as raw material that is difficult to extract, and which also requires the right kind of refining. As is often the case, there is solid ground of truth in this cliché image. Many companies, organizations and research institutions face the same challenge: how to extract and process data in a cost-efficient manner that delivers business value at the end of the journey? We compare the classic project approach for AI with Trendskout AI software.