Practical Frameworks for Industrial AI, by Dr. Jack Saunders

Abstract:
Artificial intelligence is now embedded in everyday conversation, but turning that enthusiasm into working systems in real organisations is still surprisingly hard. In this talk I will outline a systematic, repeatable approach to starting and scaling AI projects in industry, drawing on case studies from real-world projects we have undertaken in industries such as water utilities, education and healthcare. I will sketch a simple set of questions that any proposed use case must answer – clarifying the decision we are trying to support, the data we can realistically access, how success will be measured, and the constraints of the environment the system will operate in – and use this as a thread we return to throughout.
Examples will include prioritising leakage repairs in water networks, connecting school staff to policy documents using retrieval-augmented generation, and building secure, patient facing tools for healthcare providers. In each case, we will use the same questions to examine data quality, evaluation design, uncertainty, bias, and crucially whether the model’s performance is good enough to justify changing existing processes.
The aim is not to showcase exotic architectures, but to highlight where a structured approach makes the difference between a clever prototype and a tool that genuinely delivers impact. Attendees should come away with a practical model they can apply to stress-test AI ideas in their own organisations, and to collaborate more effectively.

Jack Saunders is Technical Director at Trident Intelligent Solutions, a data science and AI consultancy that designs and deploys systems in complex, highly regulated environments. He leads multidisciplinary teams delivering projects across a variety of sectors, with a focus on turning promising ideas into robust, production-grade tools.

Holding a PhD in machine learning, Jack brings a research-driven perspective to industry challenges, ensuring that cutting-edge methods are applied responsibly and effectively. He is committed to building systems that are not only innovative but also dependable, secure, and aligned to the operational needs of the organisations they serve.