AI & Data Best Practices
The Five Steps for an AI Project: What you’re missing
You might be looking for an approach to use when running your AI project. Perhaps you did a search for AI project management methods or ways to manage an AI project. In doing so, you might have stumbled across some references to a Five Step process for running an AI project. Where did this come from? And more importantly, when it comes to the five steps for an AI project, what are you missing?

Join Thousands of Others Who are Certified in AI Best-Practices
You might be looking for an approach to use when running your AI project. Perhaps you did a search for AI project management methods or ways to manage an AI project. In doing so, you might have stumbled across some references to a Five Step process for running an AI project. Where did this come from? And more importantly, when it comes to the five steps for an AI project, what are you missing?
A few years ago, well-known AI researcher and entrepreneur Andrew Ng developed a high-level, simplified general approach to running AI projects that consists of a five step framework to plan AI projects effectively. The five steps were put out more as a high-level concept for how to run AI projects than a specific, detailed plan or methodology for running those projects. Regardless, those five simplistic steps for an AI project have been making rounds across the technology industry.
While the five steps for an AI Project mentioned are important aspects of successfully implementing AI projects, the specifics leave much to be desired. In fact, more is missing in the five steps for an AI project than is present in the original post about the approach. Let’s dive into the details and see what’s missing from the five steps for Artificial intelligence project management.
What are the Five Steps of an AI project cycle?
The five steps outlined in Andrew Ng’s AI project management article are:
- Identify a business problem (not an AI problem)
- Brainstorm AI solutions
- Assess the feasibility and value of potential solutions
- Determine milestones
- Budget for resources
These five steps make a good amount of sense, and in many ways are fairly obvious. Indeed, we need to identify a problem, figure out a solution, make sure the solution is feasible, determine milestones, and then budget for resources. But aren’t we forgetting a bunch of things? Like when do you actually implement the AI project? And what about data preparation, augmentation, and collection which often take up 80% or more of a project? How exactly do you brainstorm and evaluate AI solutions? Where do you evaluate machine learning models to make sure they actually work? And when and how do you deploy and operationalize AI systems? None of these details are included in that very high-level five steps of an AI project plan.
The Five Steps for an AI Project are Really just the First Step in a Best Practice Methodology (CPMAI)
One thing that is immediately evident about these Five Steps for an AI Project lifecycle is that they are all just the first step in an iterative, best-practices project management methodology. One of the best practices for AI project management is the Cognitive Project Management for AI (CPMAI). CPMAI is built upon the well-established, data centric CRISP-DM, and incorporates best-practices iterative approaches for short sprints and serves as a solid AI Project guideline to run your AI and machine learning projects. CPMAI is the fastest growing AI project management course and certification for running and managing AI projects, with 220% annual growth rate.
The CPMAI AI project management methodology for AI projects is divided into six iterative stages of AI development that ensure you address the most important considerations for an AI project at the right time to optimize for success as part of the AI development process. The six CPMAI steps, or phases are detailed below:
Source: Cognilytica
Description: The Steps for an AI project follow six main phases, from the CPMAI methodology
CPMAI Phase I: Business Understanding
The first phase in any AI project is gathering an understanding of the business or organizational requirements. Borrowing from CRISP-DM, but customizing for AI purposes, the CPMAI Phase I Business Understanding phase focuses on understanding the project objectives and requirements from a business perspective, then converting this knowledge into an AI and cognitive project problem definition and a preliminary plan designed to achieve the objectives. CPMAI Phase I Business Understanding aims to address three key needs for AI projects: Business Requirements, determining which (if any) pattern or patterns of AI meet those business requirements, and identification of what the most important deliverable would be for that particular iteration of the AI Project.
CPMAI Phase II: Data Understanding
Once you have a solid business understanding of the context and value of the AI project, the next step is CPMAI Phase II Data Understanding. Since AI projects are at their core data projects, a firm understanding of the data environment and availability is absolutely essential to AI project success. In the Data Understanding phase of the AI project iteration, you should look to address the crucial data requirements for AI projects: the availability and sources of data to meet business needs, the quality of that data and need for enhancement or augmentation, and the environments in which data is needed for training and real-world inference.
CPMAI Phase III: Data Preparation
With business and data requirements identified, it’s time to figure out what you need to do with that data to get it into a “model ready” state. CPMAI Phase III Data Preparation includes, data cleansing, data aggregation, data augmentation, data labeling, data normalization, data transformation and any other activities for data of structured, unstructured, and semi-structured nature. It’s important to wrangle data from the sources and transform it to its required state, data cleansing to eliminate critical data flaws, and data augmentation and enhancement including data labeling to add necessary meaning and context to the data so that you produce an accurate AI model.
CPMAI Phase IV: Model Development
The next step you need to consider focuses on aspects of model development. With data prepared for the right solution, Model Development can focus on selecting the right approaches and algorithms for the model, based on the business requirements, data availability, and performance needs, and also perform the actions of tuning and configuring the model for optimal performance with hyperparameter tuning, determining if you will use third-party models and/or extend those models, matching the model performance against business requirements as well as perform necessary model training activities.
CPMAI Phase V: Model Evaluation
Model evaluation is essential when it comes to ensuring that an AI project methodology delivers the expected value. How will you know if your model is meeting your business needs if you don’t test it? CPMAI Phase V Model evaluation focuses an AI project on model metric evaluation, model precision and accuracy, determination of false positive and negative rates, key performance indicator metrics, model performance metrics, model quality measurements, and a determination as to whether or not the model is suitable for meeting the goals of that iteration. This is an aspect of project management for AI & data science that can be easily overlooked. If you don’t fully evaluate your model you could very well end up deploying a model that will provide you with no real value.
CPMAI Phase VI: Model Operationalization
With a proper model developed and evaluated that meets business requirements using the data available and prepared, the model can be put into a production, or operational, environment with some degree of success. In CPMAI Phase VI we focus on putting AI models into the real-world. However, the responsibility doesn’t stop there. AI projects aren’t “set it and forget it” projects. The real world changes, which changes the data, which means if you aren’t continually iterating on your model it will decrease in value over time. To be successful, you must continually iterate, update and test your model to make sure you don’t run into issues around data or model drift and to make sure your model is continuing to provide value. Model operationalization best practices include model versioning and iteration, model deployment, model monitoring, model staging in development and production environments, and other aspects of getting the model in a position to provide value to meet the stated purpose.
What are the missing steps for an artificial intelligence project plan?
It’s fairly clear that the five steps of Andrew Ng’s Framework fall into just the first step of the CPMAI AI project management certification and methodology, CPMAI Phase I Business Understanding. The business understanding step focuses on understanding the project objectives and requirements from a business perspective, identify which patterns of AI you’re addressing, identifying which parts of the problem should be solved in a non-AI way, key performance metrics and measures, ethical guidelines and boundaries, ROI and budget requirements and provides an iterative artificial intelligence project plan designed to achieve the objectives.
Furthermore, to be actually successful with an AI project, you can’t just focus on the initial business requirements and brainstorming. You need to focus on the harder aspects of data understanding and preparation, model development and evaluation, and model deployment and iteration. WIthout taking these into account, those great AI ideas you brainstormed will never become a successful reality.
CPMAI certification provides the needed AI project management certification organizations and individuals are looking for to achieve the success and get the detail that the simple five-step approach doesn’t provide. The CPMAI training and certification helps organizations take their AI project management success to the next level. By getting your CPMAI certification you are joining the fastest growing certification for Artificial Intelligence project management. In addition to the training and certification, CPMAI provides an AI project template and workbook that can be applied for every ML and AI project you plan to run. Download our CPMAI project checklist to give you guidance on the steps for AI project success.
You can also dig deeper into the various AI and data concepts in our AI & data resource and reading list.
Learn more about CPMAI and join the thousands of certified data and AI project management professionals who have implemented CPMAI across both public and private sectors and are delivering successful AI projects. You can sign up for CPMAI certification on our site and get started today!