April 28, 2017 at 10:25 AM
In this blog post, I will be covering some of the highlights from day 2 of the Finovate Spring Conference. Whereas the previous review of day 1 mostly covered innovations that assist the loans origination process, in this blog I’ll cover some of the analytical offerings. I’ll also cover some of the other offerings including training, investment platforms, and bot-based technology.
Analytical based innovation at Finovate
Throughout the conference there were really four stand-out companies in the data analytics space.
1. MapD streamline Big Data BI
The first was the Big Data BI tool offered by MapD. The difference between MapD and some of the other BI tools that I have seen was their use of graphics processing unit (GPU) to process billions of rows of data. On top of that it features some of the standard slice-and-dice, mapping and drill-down capability.
2. Neener Analytics use social media and linguistics to minimise credit risk
Over the 2 days of presentations there were a few companies that touched on using social media data in credit decisions. Whereas social media data has traditionally been used for thin-file assessment by making the large assumption that “you are like your friends” or “you are like those who like the same brands as you”, other approaches were also demonstrated. Neener Analytics have brought together professors (including a Nobel Prize winner) whose expertise cover social media linguistic expertise essentially looking at how language is used in an applicant’s social media posts. Through structuring and aggregating this data they report that they are able to predict FICO score ranges, defaulters, transactors/revolvers and early settlers.
3. AlphaRank predict churn through analysis of transactional data
The third analytics organisation of interest was AlphaRank. Whereas traditional anti-churn strategies involve using predictive analytics looking at customer behaviour to determine whether a customer is likely to attrite and these models are relatively predictive, the actions taken thereafter can be of limited success. They claim that this is due to the fact that once attrition is predicted to happen, the customer has already made up their mind that they are to leave. Their solution involves building links between customers through analysing transactional data (credit card and mobile data was mentioned). From here “influencers” are identified. When a customer leaves the organisation, the organisation should take anti-churn actions on those to whom the customer is most connected.
4. AccountScore determine risk scores using transactional data
And finally, the fourth organisation that has realised value through aggregating transactional data is the UK-based AccountScore. They have built transactional data based risk scores that have been adopted by banks in the UK and USA.
Some of the worthy mentions were some of the investment apps. Quantiacs offer a platform for analysts to import large files of time-series stock market data. From here a bunch of analysis can be conducted and the analyst can build their own trading algorithm through a parameterised rule-building tool. This trading program can be run retrospectively to determine the return. The trading tool can then be compiled automatically in Python and placed in the marketplace. Investors can browse the trading programmes and invest. The analyst (and Quantiacs) earn commission on the investment. Essentially this marketplace democratises the hedge fund market.
NCR demonstrated their ability to build virtual-reality (VR) based training tools enabling a fully immersive and collaborative training environment. Below you see a photo taken from their demo at Finovate of hands-on remote staff training using virtual reality.
This was the 10th edition of Finovate’s Silicon Valley event. Like our last conference two years ago, it was a highly recommended conference with a lot of new ideas. For those who can’t make it, all presentations are uploaded to their website. When these become available we will link to them below.