Candidates for this certification have a foundational knowledge the procedures used to develop an
artificial intelligence (AI) solution, as well as an understanding of the issues surrounding the governance,
transparency, security, and ethics of AI. Successful candidates will be able to analyze and classify a
problem. They should be able to demonstrate knowledge of data collection, data processing, and feature
engineering strategies.
1. AI Problem Definition
1.1 Identify the problem you are trying to solve using AI
(e.g., user segmentation, improving customer service)
1.2 Identify the areas of expertise and resource requirements needed to
solve the problem
1.3 Classify the problem (e.g., regression, unsupervised learning)
1.4 Plan for AI to be used responsibly
1.5 Choose transparency and validation activities
2. Data Collection and Transformation
2.1 Identify data sources and requirements
2.2 Assess data quality
2.3 Convert data into suitable formats (e.g., numerical, image, time series)
3. Feature Engineering
3.1 Select features for the AI model
3.2 Engage in feature engineering
3.3 Identify training and test datasets
3.4 Document data decisions
4. AI Algorithms and Models
4.1 Consider applicability of specific algorithms
4.2 Train a model using the selected algorithms
4.3 Evaluate model accuracy, precision, and sensitivity
4.4 Identify potential sources of bias
4.5 Select specific model after experimentation
4.6 Tell data stories and obtain stakeholder approval
5. Application Integrity and Deployment
5.1 Design a production pipeline, including application integration
5.2 Identify potential challenges of models in production and plan for
5.3 Build a security plan
5.4 Train customers on how to use product and what to expect from it
6. Maintaining and Monitoring AI in Production
6.1 Engage in oversight
6.2 Assess business impact (key performance indicators)
6.3 Measure impacts on individuals and communities
6.4 Handle feedback from users
6.5 Evaluate operation on a regular basis





