Name:
BTN:
Fitria Zahroh, Decision Management & Analytics Head
Indra Hidayatullah, Data Management & Analytics Division Head
Joko Christianto, IT Strategic Planning & Development Division Head
Andreas Fredrik, Data Scientist
Antonius Hermawan, Data Management Department Head
Budi Muchdarli, Software and Database Staff
Aderian Primaraka, IT Project & Portfolio Department Head
Willy Kusuma, IT Business Partner Specialist
Bayu Primantio, Application Support Engineer
Izeno:
David , Software Engineer
Natalia Kemal, Senior Project Manager
Irida Wanti Sugito, Enterprise Account Manager
Deddy Johari, Director, ASEAN, Chief Enterprise Architect (Data, Analytics, Cloud, and DevOps)
Country: Indonesia
Organization: PT Bank Tabungan Negara, Tbk (Persero)
Bank BTN (PT Bank Tabungan Negara, Tbk [Persero]) was founded in 1897 with a company size of more than 10,001 employees. It is a state-owned enterprise bank, responsible for providing financial services to the public in order to drive economic growth for Indonesia in general and for small-medium businesses in the country.
Awards Categories:
- Most Impactful Transformation Story
Business Challenge:
With its main business in mortgage loans, BTN has the vision to be Southeast Asia’s best mortgage bank in 2025. To achieve this, BTN has begun the transformation process, with one of our strategies involving building a digital ecosystem. The Data Management Division in BTN plays an important role in this transformation to support data-driven decisions within the organization. However, there are some challenges that we are facing specifically in terms of process and people related to data management.
One of them is limited resources for data scientists, engineers, and analysts. Other than that, scattered data sources and different tools used for data processing (SQL Based Tools, SAS, R, Python) greatly impact the number of data-driven use cases delivered. We take too much time doing data Integration with multiple data marts and data sources and analyzing data to understand it better (visual, descriptive analytics). Data preparation is also one of the pain points we face before training the data with different algorithms and consolidating internal models built with different platforms.
Because the above processes are done with various tools, we need a solution that can complete all of them on a single platform, preferably one less focused on code. It will reduce the time that we need to deliver our model solution.
Business Solution:
When we were looking for a solution, we wanted to have a data science platform that could not only perform machine learning solutions but also complete machine learning processes in one single platform. We wanted it to be up-to-date and able to accommodate the newest machine learning algorithm updates from another platform. Moreover, the tool needed to be easy to use, even for people with minimal machine learning knowledge and coding skills. We also wanted a tool that could perform data integration from multiple sources with minimum effort, visual and statistical analytics, and create a dashboard to present insight from the data.
After six months of using Dataiku, our team is able to perform multiple end-to-end machine learning processes with significantly reduced time. Dataiku was also able to provide solutions to our difficulties in integrating multiple data sources and performing data transformation with the minimum effort that we usually require when using our ETL platform.
We're also able to import updated resources with ease, and integrated our internal model within the Dataiku platform to make the coordination of team projects easier. One model which would typically take three months can now be done in less than a week.
The model we applied had a significant impact on our campaign. We can deliver insights and suggestions to the business team regarding what campaign to focus on. What's more, we discovered a potential portfolio incremental of more than 1 Billion USD in our first three months using Dataiku.
Value Generated:
In addition to optimizing our resources and improving our data management process, we can improve our data team coordinator using the Dataiku platform. It helps teams as well as the manager track and distributes the job between each person.
With all of these advantages, we can deliver more use cases in a shorter timeline. One model which would typically take three months can now be done in less than a week.
Value Brought by Dataiku:
The fact that it's a single end-to-end integrated data science platform like we were searching for, for strong support in making seamless data integration.
The Dataiku platform brings a new level of ease to data preparation and model development. With this edge, We developed the Segmentation Model & Next Product To Buy in Dataiku and collaborated with our business team on the action plan.
The model we applied had a significant impact on our campaign. We can now deliver insights and suggestions to the business team and discovered a potential portfolio incremental of more than 1 Billion USD.
