Alfa, provider of Alfa Systems, released its second paper on artificial intelligence in the industry. Part 2: Using Machine Learning in the Wild is a more technical follow-up to 2019’s Part 1: Balancing Risk and Reward and explores in two specific use cases which take very different approaches to machine learning implementation. It features a foreword from Blaise Thomson, whose speech technology startup VocalIQ was acquired by Apple and formed a part of the Siri development team.

“AI and machine learning are front and center in the asset finance conversation at the moment but many don’t know where to start — how much expertise they need, what they can outsource, and where they should concentrate their efforts and costs,” Martyn Tamerlane, a solution architect at Alfa and co-author of the paper, said. “Our worked-through examples convey genuinely useful and practically applicable advice for people wanting to kick off their own machine learning projects. By comparing the approaches used, we offer advice on what’s right for others.”

The first example, which addresses automated license plate recognition and its ongoing embedding in business processes, takes an off-the-shelf approach to training machine learning models, drawing heavily on tools provided by AWS. Meanwhile, the second, which analyses Alfa’s internal code tests, is carried out wholly in-house with existing resources and knowledge. The paper also features a decision aid to help readers clarify how their projects might compare.

2019’s Balancing Risk and Reward outlined the high-risk, high-reward nature of using AI in the asset finance industry, and machine learning in particular. Alfa will continue its commentary on AI in asset finance with further upcoming publications.