AI Engineering: Twelve Foundational Practices
• Brochure
This update to the 2019 AI Engineering: 11 Foundational Practices provides timely recommendations for decision makers, including promoting evaluation to a primary engineering concern.
Publisher
Software Engineering Institute
Topic or Tag
Abstract
The core ideas of the SEI's AI Engineering: 11 Foundational Practices have remained relevant since its publication in 2019, but the practice of building, using, and deploying AI has changed significantly and continues to evolve in 2026. In this update, evaluation is promoted to a core engineering practice. Data considerations expand across all phases of the system lifecycle. Security, traceability, uncertainty, oversight, and modularity are all revisited in the context of foundation models and agentic systems.