Artificial Intelligence Engineering Body of Knowledge
Created October 2022
AI engineering is an emergent field of research and practice that combines the principles of software engineering, systems, computer science, and human-centered design to enable implementation of AI systems in accordance with human needs for mission outcomes. While much work to date has focused on AI capability development, AI engineering asks what is needed to develop, deploy, and maintain AI systems in unpredictable and chaotic environments.
In working to grow this discipline, the SEI has been talking with stakeholders about the challenges they face when implementing AI systems. We are looking towards the broad and diverse community and mission partners to help develop an AI Engineering Body of Knowledge that will equip practitioners with the knowledge and skills to move systems from research into practice.
Developing the AIEBoK
The AI Engineering Body of Knowledge will be a common vernacular for diverse stakeholders to establish common principles and practices to deliver reliable, responsible, and effective solutions. It will guide research and leverage repeatability, reusability, and modularity in system design. The foundations of this discipline are the three pillars: Human-Centered AI, Robust and Secure AI, and Scalable AI.
Using these pillars assures that systems are designed with and for people, that they operate reliably when faced with uncertainty or threat, and that they accommodate size, speed, and complexity of mission needs to be reusable and mission ready.
The vision includes collaboration on an open-source, GitBook-based project for developing and maintaining the shared knowledge. It will complement existing frameworks, such as the Software Engineering Body of Knowledge and the Systems Body of Knowledge, will involve community support and peer reviews, and will standardize practices and expectations across technical, cultural, and managerial tenets in a dynamic framework.
Who Benefits?
The ability to reference and refine an AIEBoK benefits everyone within the AI community and beyond. It will give stakeholders and end users a trusted resource to access information and gain more understanding of the systems. Stakeholders include those in the thick of the development of systems; software engineers, data scientists, and data engineers; the product managers who supervise the structure of the work; and every other type of individual who accesses the data.
Looking Ahead
Through community co-development we will design, develop, release, and maintain the AI Engineering Body of Knowledge. We are seeking participants across government, academia, and industry to influence the BoK as it becomes a resource to apply AI to real-world scenarios.