AnnouncementIntroducing MongoDB 8.0, the fastest MongoDB ever! Read more >>Introducing MongoDB 8.0, the fastest MongoDB ever! >>

Indonesia’s digital bank pioneer creates more smiles with the help of MongoDB Atlas

Amar/MongoDB Hero

INDUSTRY

Financial Services

PRODUCT

MongoDB Atlas

USE CASE

Cloud Data Strategy

CUSTOMER SINCE

2014
INTRODUCTION

Changing lives through data-driven technology

To Bring 200 million smiles by 2025 – that’s the goal of Amar Bank.

Since 2014, the bank has been a pioneer in digital finance solutions that positively impacts people’s lives in Indonesia. It provides innovative and safe ways to plan and manage finances through leading-edge technology.

Amar Bank was the pioneer of digital banking in Indonesia to focus on providing microloans to people who wouldn’t be able to get financial services from traditional banks (unbanked and underserved). For many, typical hurdles to accessing smaller loans include not having a credit history or the collateral to secure a loan. In addition, conventional lenders have a daunting paper application process that takes much time and effort to complete.

Given the obstacles many potential borrowers face, Amar Bank created the AI-led, intelligent Tunaiku, digital lending platform, to enable more people to get the finance they need quickly and easily without the traditional lending requirements. Customers complete a simple application on the Tunaiku apps, get a quick assessment, and can be accepted for a loan within minutes.

By leveraging technology and data, Amar Bank has been able to break down the barriers so their customers can access small loans to make a big difference for themselves and their communities.

For example, a small business owner running a laundromat might have several washing machines out of service but lacks the funds to get them repaired. This means less income for his family and community. In this case, getting a small loan from a conventional bank would not be feasible, but Amar Bank through Tunaiku, is able to help.

THE CHALLENGE

How to use non-structured data to make lending decisions quickly and efficiently

Seeking alternatives to the traditional credit decision model created several challenges. While traditional underwriting processes have proven effective in determining the creditworthiness of potential borrowers, they would not work for the market seeking smaller loans to people who don’t have a credit history or collateral. Lending small amounts, up to around US$1,500, and having tight margins also meant that the business would rely on a large volume of loans to be profitable. So solutions that would scale efficiently, reliably and cost-effectively were the key.

While innovative approaches to managing and analyzing data were the key to the solution, it raised major questions for Amar Bank, including:

  • How can we assess and manage data to make effective lending decisions?
  • How can we manage a wide variety of data?
  • How can we scale quickly?
  • Is there a flexible data solution or platform to cover usage spikes at busy times?
To develop a decision model based on predictive analytics the team uses some structured data (such as names and addresses) and a lot of unstructured data to determine the creditworthiness of potential borrowers. The unstructured data includes many situational and behavioral factors. A solution was needed that could connect both sets of data.

THE SOLUTION

Amar Bank overcomes data challenges and scales with MongoDB

When Amar Bank started operating in 2014, the data was on-premises. After a few months, they started migrating data to the cloud. Also in 2014, Amar Bank started using MongoDB Community to manage the massive amounts of non-relational data. MongoDB gathers the complex situational, and behavioral data used to make lending decisions. This data sits alongside the structured data. A key factor in selecting MongoDB, is that it’s a document-oriented database, making it easier to store and query the two types of datasets.

MongoDB has played a critical role in creating and implementing its credit model. As Kevin Kane, Chief Technology Officer at Amar Bank, explained, “To develop our credit model, we created a single line between structured and unstructured data to conduct data modeling using MongoDB. When we prove that the model works, we put it into production. Once deployed to production, we use both types of data to provide a score determining if we offer a loan and the loan amount.”

The loan approval process takes 24 hours, which gives the bank an edge in serving customers who need a fast financial solution.

One of the reasons Amar Bank started using MongoDB at an early stage was that they didn’t know the scope of the data points needed to develop and apply the decision model and required a flexible solution.

Kevin values this flexibility and said, “As the scale increases and you need to use more data points, MongoDB can help you grow and scale to manage new types of unstructured data.”

The data needs to grow with the business and Amar Bank has continuously adapted. As Ahmad Fikri, VP Infrastructure, Operation & Cyber Security, noted, “We loved using the open-source version of MongoDB, but we needed something super seamless. As the company started to grow and add more customers and data, the operation and management of data became more challenging.”

Amar Bank began using Google Big Query for analytics, which was the starting point of moving everything to Google Cloud. After they migrated their final application to Google Cloud in 2020, they also decided to upgrade to MongoDB Atlas to assist with operational management and support more flexible scalability. Since MongoDB has a close partnership with Google, Atlas runs natively on Google.

One big plus of moving to MongoDB Atlas was the seamless migration from self-managed community to the fully managed Atlas service.

“The migration to MongoDB Atlas was very simple with absolutely no downtime. With some of the other database technologies, we had some issues getting it into the cloud,” Ahmad said.

Another benefit with MongoDB Atlas is auto-scaling which automatically increases capacity as the number of customers and usage increase and dials down when customer use decreases. This helps keep costs low in a business where profit margins from lending are tight.

“The migration to MongoDB Atlas was very simple with absolutely no downtime. With some of the other database technologies, we had some issues getting it into the cloud”

Ahmad Fikri, VP Infrastructure, Operation & Cyber Security, Amar Bank

THE RESULTS

Delivering hundreds of millions of dollars to Indonesian economy

Since its establishment in 2014, Amar Bank has gone from strength to strength. It has grown its employees from a total of 17 people in 2014 to more than 1,080 in 2022 and has brought home 28 prestigious awards since 2017. These included the TOP Digital Implementation Award and Asia Pacific Enterprise Awards (APEA) in 2021.

As a result of making credit accessible to more people, Amar Bank disbursed USD $537 million (IDR8 trillion) in small loans to more than 575,000 customers in 2022. This resulted in a huge increase in revenue in seven years – from USD $1.07 million annually to USD $48.5 million.

The next step for Amar Bank is extending its innovative technology to the MSME space. So businesses without financial statements and credit histories that conventional lenders require, can still get life changing loans. MongoDB will play a crucial role in managing new types of unstructured data to create and apply a business-focused credit decision model.

MongoDB also keeps the options open for going multi-cloud. “MongoDB is cloud-agnostic. The multi-cloud element gives us flexibility as we grow and adapt our cloud strategy. MongoDB opens the door for our multi-cloud future,” Kevin said.

The multi-cloud element gives us flexibility as we grow and adapt our cloud strategy. MongoDB opens the door for our multi-cloud future”

Kevin Kane, Chief Technology Officer, Amar Bank

What will your story be?

MongoDB will help you find the best solution.