Identify Data Needs

In the PMI Certified Professional in Managing AI (PMI-CPMAI) framework, 'Identify Data Needs' is a core domain that corresponds largely to the Data Understanding phase of the CPMAI methodology. CPMAI is an iterative, data-centric approach to running AI projects. It stresses that AI projects succeed or fail mainly because of data, not algorithms. After the business problem and the AI pattern have been defined (for example recognition, prediction, conversational, or anomaly detection), the project manager must determine what data is required to solve that problem and whether that data is actually available and usable. Key activities include the following: 1. Translating business requirements into data requirements. This means defining the features, labels, volume, granularity, and time range the model will need. 2. Identifying and locating data sources. These can be internal systems, third-party providers, public datasets, or synthetic data. Ownership and access rights should be documented for each source. 3. Assessing data quality and quantity. The team examines the 'V's of data (volume, variety, velocity, veracity) and checks completeness, accuracy, consistency, timeliness, and representativeness. 4. Determining labeling and annotation needs. For supervised learning, the team must decide how ground truth will be created and validated, and at what cost. 5. Evaluating governance, privacy, security, and compliance. This includes regulations such as GDPR, consent, intellectual property, and data residency. 6. Identifying bias and ethical risks. Representative and fair data supports trustworthy AI. 7. Planning infrastructure and pipelines. The team defines how data will be collected, stored, moved, and versioned. A critical outcome is a go/no-go decision. If the needed data is unavailable, of poor quality, or too costly to obtain, the team should iterate back to business understanding and rescope the project rather than continue. For the exam, remember a few principles. Data feasibility is assessed early. The process is iterative, not linear. Collaboration with data owners, data engineers, and stakeholders is essential. Every decision must tie data back to measurable business objectives and responsible AI principles. In the PMI Certified Professional in Managing AI (PMI-CPMAI) framework, 'Identify Data Needs' is a core domain that corresponds largely to the Data Understanding phase of the CPMAI methodology. CPMAI is an iterative, data-centric approach to running AI projects. It stresses that AI projects succeed…