Francis Flannery is a Principal Solutions Architect with Amazon Web Service where he helps companies of all sizes build on AWS. Francis has worked in AI/ML for well over fifteen years. One of his focus areas is bringing generative AI use cases into production. He has a degree in Electronic Engineering and a Research Masters in “Automating the formal Verification of Security Protocols”. Previous to AWS he worked at the University of Limerick, Intel, and a number of start-ups including his own.

Description:
With the advent of generative AI the ability to build proof of concepts has never been easier while the path to production continues to be challenging. The move from AI/ML models from being a backend process to now often becoming the de-facto user interface leads to challenges around how to be safe, secure, and cost effective. The level of innovation and change in the field is conversely leading to organisations unsure of when or how they should scale their generative AI use cases in production. All of these factors are leading to organisations unable to move from proof of concept mentality into production i.e. being in POC purgatory. This talk will cover the multi-faceted challenges organisations need to overcome, discuss some successful strategies seen in the field, key trends, and areas of continued research.