Canadian Underwriter
News

The ‘biggest hurdle’ to making artificial intelligence useful


November 28, 2018   by Greg Meckbach


Print this page

Figuring out what products your customers might buy next or what they are complaining about are two ways you could be using artificial intelligence now, the founder of a big data vendor suggests.

“In the future you are going to see more technology organizations that just happen to be selling shoes or just happen to be selling insurance but first and foremost they want to present themselves as a technology company,” Clara Angotti, president and founder of Next Pathway Inc., said Wednesday in an interview.

Some companies are using AI to order to understand what customers are purchasing to predict what they are going to purchase or at what price point they will make that purchase.

“In simple terms AI is about incorporating human intelligence into machines,” said Angotti, recently named one of Canada’s 100 most powerful women this week by the Women’s Executive Network.

Angotti cited speech recognition, image interpretation and language translation as examples of how AI is used today.

There is already artificial intelligence software on the market, but AI is more than just the algorithms that pore through data.

There is gathering the data itself, and data comes in different formats – such as Twitter feeds, Facebook likes, images and voice recordings, Angotti noted.

“The question someone might have is, ‘what are my top 5 complaints? Of those top 5 complaints, which of those are coming from my customers that spend the most money on my product?’”

There are AI tools for this but “ingesting all that data to make it meaningful – that becomes the biggest hurdle for people,” she said.

“Putting all those things together in one database and then transforming it to a common data model – this is very technically challenging for most organizations.”

Next Pathway’s core business is big data. The vendor puts massive amounts of data into a “data pipeline” and standardizes file formats so the user can attach an algorithm to it in order to interpret the data and build a predictive model.

With its data pipeline, Next Pathway says it can help customers prepare their data more quickly – in some cases, weeks or days instead of months.

One of the biggest challenges with AI is “to take multiple files that come in in different file formats and be able to standardize them so that you can ingest all the data that you want and be able to analyze it across multiple data sets,” Angotti said.

Some insurance providers, banks, financial services firms and retailers are doing a good job at using AI to understand purchasing behaviour and predicting what customers will purchase next, she added.

“Those companies that take advantage of technology and take advantage of the data that is already resident in their own organizations will become the winners.”

WXN announced Nov. 22 its 2018 Canada’s Most Powerful Women: Top 100 Award Winners. Angotti received the award this year and in 2011.

Her “ability to anticipate market trends and deliver with quality and excellence has allowed her to build and sell two technology businesses,” WXN said.

The 100 winners are “proven achievers who are strong contributors to their organizations, their fields of endeavour and their communities, says WXN.