A data pipeline moves raw data through collection, cleaning, processing, and presentation so analysis can support business decisions. Its shape depends on the data, question, methods, and audience. section 1.2.1
A typical analytical pipeline:
- Collect and clean data from relevant sources. section 1.2.1
- Standardize the data and process it with appropriate algorithms. section 1.2.1
- Present results to the audience who must use them. section 1.2.1
Different questions, different pipelines
A retail pipeline can combine transactions, customer feedback, and website interactions for segmentation, product affinity analysis, and sales forecasting. Healthcare and manufacturing pipelines use different inputs and analytical goals. section 1.2.1
Figure 1.1 Recommended generative AI–supported data analysis flow
Use the flow diagram to trace how generative AI can support the stages of data analysis. section 1.2.1
