Survey banner
Switching to Dataiku - a new area to help users who are transitioning from other tools and diving into Dataiku! CHECK IT OUT

Davivienda - Developing a Financial Health Indicator (ISF) to Better Understand Customers and Improve Collections Strategies

Name:

Andres Hoyos
Esteban Ortiz

Country: Colombia

Organization: Davivienda

Davivienda is a regional bank leader that operates in six countries (Colombia, Panama, Costa Rica, El Salvador, Honduras, and the United States). With over 50 years of experience in the Colombian market, Davivienda offers a wide range of financial services to individuals, SMEs, and corporate customers. The bank currently serves over 20 million customers through a network of more than 650 branches and 2.500 ATMs.

 

Awards Categories:

  • Best Acceleration Use Case
  • Best MLOps Use Case
  • Best Moonshot Use Case

 

Business Challenge:

As a credit institution, the Bank constantly faces the challenge of approving credits optimally by implementing various techniques and policies to mitigate the risk of default. However, there are situations where customers may default due to various factors, resulting in losses and risk costs for the institution.

Therefore, it is crucial to find an efficient and solid way to focus the efforts of the collections department in effectively managing the portfolio. This involves defining more aggressive strategies for customers who pose a greater challenge in terms of payment.

The objective is to find the right balance between maximizing recovery, reducing operational costs, maintaining a good relationship, and appropriate service levels with customers.

To address this challenge, the development of a financial health ecosystem was proposed, which allows an understanding of the life stage of users and improves collection strategies. One of the tools designed within this ecosystem is a probability of payment score, which was constructed using segmentations and machine learning techniques, taking into account the customer's historical information, obligations, and behavior in the financial sector. This score, along with different policies and considerations of the institution, enables the definition of a more robust collections strategy tailored to each customer.

 

Business Solution:

Dataiku played a pivotal role in addressing this challenge by facilitating the development, implementation, and production deployment of the required model and segmentation. Its platform provided an intuitive interface that allowed us to access and manipulate large volumes of customer information, totaling approximately 20 million records.

Our approach focused on conducting two stages of segmentation prior to model development, aiming for maximum specificity and delivering more personalized treatment to our clients. This led to the creation of a total of 18 models, each tailored to a specific segment.

The entire process, from data preparation to final implementation, was carried out entirely within the Dataiku platform. This enabled us to have complete tracking and traceability of the process, optimizing production through the high-speed and high-quality processing capabilities offered by the tool, as well as the collaborative and efficient work of the team.

The utilization of Dataiku provided us with a significant advantage by streamlining and simplifying data analysis, manipulation, and modeling tasks, resulting in a more sophisticated and effective solution to address the challenge of portfolio management and collections in our credit institution.

 

Day-to-day Change:

In our organization, we have certainly experienced significant changes in our day-to-day operations. By adopting new technologies and solutions like Dataiku, we have been able to improve and streamline the processing and execution times of the ISF project assessments, which would not be possible using other tools.

First and foremost, we have observed enhanced agility in our daily operations. Through the automation of routine tasks and process optimization, particularly in scenario creation, we have significantly reduced the time spent on manual execution activities. This has enabled us to provide a faster response to the collections team, empowering them to utilize the outputs and generate their respective strategies promptly.

In terms of collaboration, Dataiku has facilitated communication and teamwork within the organization. The platform has provided us with the ability to share and collaborate on projects efficiently, fostering a collaborative work environment and continuous improvement in our projects.

 

Business Area: Risk/Compliance/Legal/Internal Audit

Use Case Stage: In Production

 

Value Generated:

The Financial Health project has generated significant value for the Collections Department by enabling us to anticipate the needs and behaviors of our customers. This use case has provided us with invaluable insights to enhance and optimize the ecosystem for debt management within the bank.

Our primary objective was to efficiently and strategically focus our collection efforts, particularly toward delinquent customers. Given our limited resources, such as budgets allocated to specialized collection agencies and other management channels, it became essential to have a tool that would allow us to prioritize our efforts and maximize results. Through customer identification and segmentation, we have been able to identify population groups requiring varying levels of attention, thereby necessitating a differentiated treatment within our debt management strategy.

 

Value Brought by Dataiku:

Dataiku has provided us with great value in multiple ways. Its ability to process large volumes of information has significantly accelerated our processes, resulting in a notable improvement in the operational efficiency of our organization.

Furthermore, the Dataiku platform has provided us with the necessary flexibility and agility to experiment and test different approaches and methodologies. This has allowed us to conduct deeper and more detailed analyses of our data, leading us to develop more efficient solutions in terms of performance and predictive capabilities. By being able to incorporate and compare different models, metrics, and methodologies, we have been able to identify the optimal solution for our specific challenges.

Share:
Version history
Publication date:
05-07-2023 09:52 AM
Version history
Last update:
‎08-11-2023 11:09 AM
Updated by:
Contributors