AI GLOSSARY

Designed for Deployment

Designing for deployment refers to an approach for developing AI systems that deliver real-world operational impact, especially in high-stakes applications. Unlike other approaches to building AI which may perform well in theory, systems that are designed for deployment must be built specifically for the problem they aim to solve. They take into consideration the environment in which they will be deployed (including the operational needs), the AI technologies that are available, and the time required to achieve success. This approach encourages iterative learning and testing with users to refine AI solutions and adapt to emerging operational requirements, making the development process a continuous cycle rather than segmented phases.

All Terms
A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z

Continue learning...

View Resources

7 min read
Machine Learning Types and Their Infrastructure Use Cases
AI and Machine learning is a complex field with numerous models and varied techniques. Understanding these different types and the problems that each...
6 min read
The Business of AI in UK Defence and National Security
While the technical aspects of an AI system are important in Defence and National Security, understanding and addressing AI business considerations...
4 min read
The 8-Step Guide to Deploying Machine Learning in Infrastructure
The adoption of AI and Machine Learning is complex, and attempting it in civil engineering can feel daunting. This guide outlines the 8 steps...

Stay connected

News, announcements, and blogs about AI in high-stakes applications.