Kreditech and Exasol

A technology platform that enables banking based on algorithms and Loading...big data scoring

Kreditech has a clear objective: to offer tailormade financial products and services to consumers who do not have a credit score. As a result, the high-growth, company has developed technology that uses Loading...Big Data and complex self-learning algorithms, and has rolled it out across nine subsidiaries around the world. Using alternative data sources – some 20,000 data points – Kreditech’s solution can decide whether a customer should be offered credit or not. Innovative in its approach, the technology enables a great degree of prediction: algorithms make decisions on issuing credit that are based on observations. Whereas traditional credit scoring models (e.g. FICO) are solely based on pre-calculated hypotheses, Kreditech’s proprietary technology takes just seconds to calculate whether credit applications are accepted or denied. To process this large amount of data volumes efficiently and to enable a fast decision-making process based on an applicant’s behavior online, Kreditech has built its innovative solution using Exasol and Tableau Software.


Challenges

Fast growth meant that the company’s database system was heavily used and analytic queries in MS Excel took up to 24 hours to run

Solutions

Exasol deployed as the central data warehouse at Kreditech, which is hosted and managed in Exasol’s private cloud, ExaCloud

Benefits

Substantially reduced maintenance and administration overhead, analytic queries run in just under two hours compared to 24 hours before Exasol

Exasol and Tableau simplify our business operations considerably, and the fast analytics we can now do have already translated into a positive return on investment.

Sebastian Ranzinger former Head of Business Intelligence at Kreditech Holding SSL GmbH
200

Staff

0

Countries

0

Data points

24 hours => 2 hours

For analytics queries to run

Background

Kreditech currently employs over 200 staff from 40 countries and has a vision of creating the “digital bank” of the future. Using its integrated financial platform solution, the company offers micro loans, instalment credits, e-wallets and other products, with further services added to its portfolio in the future. Self-learning algorithms then calculate the probability of customers paying back their loans, based on a wide range of data points. For example, one data point could be whether the applicant hesitates when asked for their employer’s details or fails to give an average income or valid place of residence, or makes a false statement which does not correlate correctly with other data captured in the application process.

However, the decision to award credit is not made on the basis of one data point alone; the decision is always made on the basis of the overall picture that is given by the applicant, who incidentally offers their data willingly. Kreditech’s technology then provides a fast, customer-friendly service online and in real-time, and ensures that human errors are avoided.

  • Industry Financial Services
  • Application Loading...Data Science
  • Database Exasol. PostgresSQL
  • BI Tool Tableau
  • ETL Tool Own application, Plain SQL Scripts
Case study background details

The journey

You can use our analytics database with our private cloud, any public cloud, on premises – or as part of a hybrid deployment.

The challenge

Every decision is based on data

Established as a start-up in 2012, the Kreditech team faced the challenge of securing the availability of the data they needed. The company had to build up a set of master data as well as a quality control system. Furthermore, the company had to ensure it complied with international financial reporting standards and regulations. “As a young company, we were running our business using open source technology and had to find a new solution fast,” comments Sebastian Ranzinger, Head of Business Intelligence of Kreditech Holding SSL GmbH. “Also, given that every type of decision at Kreditech is made based on data, we needed a sophisticated system.” A dynamic business environment and fast growth meant that the company’s database system was heavily used and analytic queries in MS Excel took up to 24 hours to run – too slow for an enterprise whose business is supplying financial services online.

The solution

Speed was the order of the day

As a data-driven business, the team around Ranzinger looked for a new solution. Primarily, the new solution had to be easy and quick to use, as well as configure and implement. The combination of Exasol as the data warehouse and Tableau as a Loading...data visualization layer on top was a solution that appealed to the team immediately. Within three days, the high performance Loading...analytic database was integrated into Kreditech’s existing inhouse application, and since then, Exasol has been deployed as the central data warehouse at Kreditech, which is hosted and managed in Exasol’s private cloud, known as ExaCloud. Using Exasol, Kreditech’s analytic queries now run in just under two hours, which is a dramatic improvement on the 24 hours they took to run previously. “Without the speed and high performance of the Exasol Loading...Loading...in-memory database, it would have been significantly more difficult to implement some of the other critical aspects of our solution,” says Sebastian Ranzinger.

The benefits for kreditec

Value-add that is much appreciated

Deployed as a cloud solution, Kreditech not only profits from a substantially reduced maintenance and administration overhead. In conclusion, “Exasol and Tableau simplify our business operations considerably, and the fast analytics we can now do have already translated into a positive return on investment; we can run ad-hoc reports at any time as well as easily create powerful visualizations and dashboards for management and marketing teams using Tableau.” The setup is complemented perfectly by a streamlined database architecture, support for the programming language “R” and round-the-clock availability of the solution, which has incidentally also led to a substantial improvement in data quality. This has been achieved through a rules-based system using LUA which detects and reports irregularities.

 

Start your Journey

Get in touch today

We’d love to hear from you if you’ve got a burning question for one of our experts – or if you simply want to talk through how to take your project forward with us, or one of our partners.

Contact Us