Towards Mature Enterprises in AI engineering – challenges and directions towards possible solutions
Link to the video: https://youtu.be/CbcLHJ3cItk
The hype of AI/ML is getting in his mature phase. A focus from a fascination about endless possibilities and business opportunities and predictions of horror scenarios, has moved to a better understanding of the real current problems, costs to solve them, risk to fail, and the real new business values.
Software Center was one of the first initiatives in the world taking the questions related how to build AI systems (aka AI-enabled systems) in an industrial context, identifying AI Engineering – engineering methods, tools, and processes focusing on system lifecycles, that are far beyond development of “the best” AI algorithms. This initiative is realized in theme 5 “AI Engineering”, and currently there are ongoing several project. Our intention is to expand AI Engineering theme with new projects.
In this seminar we will present current and future challenges in development of AI systems, in particular those that are related to Software Center partners, and that are most likely be addressed in the future projects.
We will shortly refer to a) challenges in research and practice in AI-enabled system development including management of multiple developments and operational (runtime) environments, and system quality and its verification and validation vs. ML evaluation, and b) possible directions towards solutions in shaping integrated system development process, characterizing AI components in related to the system, and ensuring integrity of the AI-enabled systems.