How do companies, such as banks, telcos, and consumer lenders tackle their customers risk assessment and avoid repayment faults?
You may quickly think of government run credit scoring systems or an individuals collection history records. While these tools are commonly used in better developed countries, they simply do not exist in many markets, and if they do, their reliability is questionable at best.
I recently had the chance to attend an inspiring presentation held by a CEO of a multi-national big data credit scoring company.
The company has developed a credit scoring engine that combines data gathered from the applicant, their social media footprint and public, open source data such as statistics or even NASA satellite imagery to assess the risk a consumer poses to the lender.
The process of assessment starts the minute an applicant opens the lenders website. Keys such as the computer make and model, the browser, the originating network, the time and weekday or even the duration it takes to complete the application and how many changes are made to data entered, all factor in to the scoring algorithms. Once enough data is presented by the applicant, their artificial intelligence starts to analyze the applicants digital social footprint. By enlisting object recognition on public pictures found of the applicant, they can determine a personas lifestyle. What are their hobbies, how do they spend their leisure time and what does their friends social environment look like? As another indicator of an individuals risk, they look at where the person lives. How many times has he moved? Does he live in a house or apartment block? Does he own his place or does he rent? To assess the surroundings of an applicants living area, they rely on crowd-sourced data and statistics, imagery and even take a birds eye view of the region from satellite imagery. Are there any parks nearby? How many benches are there in the parks? What kind of street lights are used in the area? What do public pictures of the area reveal about the quality of life?
All in all they collect and combine hundreds of data points about an individual and by doing so, can generate a credit score of any individual, anywhere in the world.
Backed by clever data science and artificial intelligence, the company has managed to develop a highly successful system of not only reducing credit fault by up to 20% but also allowing lenders to issue more loans to their consumers, in markets where it was simply not possible to create a meaningful risk assessment before this offering.
Their unique approach to addressing credit worthiness is an enlightening example of how you can generate real tangible value from massive datasets. By applying clever algorithms, advanced machine learning and artificial intelligence, and combining information from multiple sources, even to most random data becomes a valuable asset.
At Avionix we use similar approaches to handle our customers datasets and combine them in new ways, that generate new insights, allowing our customers to make informed, real-time decisions. We analyze your processes and discover data that previously has been considered of insignificant importance and by unlocking these insights, streamline and optimize your operation so you can feel confident in growing your business.