Bangladesh Financial Intelligence Unit (BFIU) sought to use a data-driven approach to generate actionable insights and strengthen the national digital finance strategy.
MSC conducted a comprehensive diagnostic study that examined usage patterns and customer behavior across demographic segments. The team used advanced analytical techniques, such as cluster segmentation and supervised learning models, to assess users. We also applied bivariate analysis to uncover correlations between DFS usage and demographic factors. MSC used behavioral analytics to generate intuitive visualizations of usage patterns and public sentiment. We defined key performance indicators, managed data quality, and synthesized insights into strategic policy recommendations.

The study generated robust evidence to refine inclusive DFS policies and provided the BFIU with targeted insights for policy action. It also laid the foundation for stronger, evidence-based decision-making to improve DFS uptake among Bangladesh’s underserved communities.

Bangladesh Financial Intelligence Unit (BFIU) commissioned the project with support from the International Finance Corporation (IFC). Optimizing loan referral services through an analysis of BRI female agents in Indonesia
Bank Rakyat Indonesia (BRI) sought to improve the performance of its female agent network to deliver loan referral services. A key component of this effort was to understand the drivers and barriers that affect agent productivity and address disparities in performance across different regions.

MSC led a comprehensive data analytics initiative that included detailed regression modeling, rigorous data cleaning, and advanced agent segmentation. The study focused on how to identify regional patterns in agent performance and pinpoint the specific factors that influence the uptake of loan referral services.We provided strategic recommendations customized to regional contexts and performance levels based on these insights.

The project delivered actionable evidence that helped BRI understand regional disparities and individual agent performance drivers. These insights enabled the design of more targeted interventions and enhanced overall agent effectiveness, improved loan referral outcomes, and strengthened BRI’s capacity to scale inclusive financial services through its female agent network.

Bank Rakyat Indonesia (BRI) commissioned this project.

 

AgriStack is India’s digital public infrastructure (DPI) designed to transform the agriculture ecosystem by establishing interoperable digital platforms that integrate farmers, government systems, and markets. At its core, AgriStack includes the Farmer Registry, Digital Crop Survey (DCS), and Metadata & Data Standards (MDDS) for agriculture. MSC helped design and implement these key registries. These systems comprise the foundational digital layer that links verified farmer identities, land records, crop production data, and market transactions that enable data interoperability and more accurate, farmer-centric decision-making.

As part of this DPI, India has already issued more than 61 million digital farmer IDs across 14 states.These IDs connect to land records, crop data, livestock ownership, and benefits, such as credit and insurance, which brings unprecedented scope and scale to digital agriculture governance.

MSC’s work conceptualizes interoperable data structures and metadata standards. We boost the government’s capability to deliver targeted interventions, enhance market access, and catalyze AgriTech innovation. The systems we helped build underpin seamless data exchange, which ranges from subsidy delivery to KCC issuance. These tools help establish the backbone for scalable, inclusive agriculture transformation across India.

The Gates Foundation commissioned this project.

In Bihar, MSC partnered with the state’s agriculture department to deliver Bihar Krishi, a flagship digital platform that now serves more than 500,000 registered farmers, with 25% being women. The app was honored with a Gold Award in the “Information Technology Initiatives for Agriculture and Farmer Empowerment” category at the prestigious ET Government DigiTech Awards 2025.

Our primary contribution was the design of a voice-first, AI-enabled interface, which featured text-to-speech advisories in local languages, voice-based search for scheme discovery, personalized soil health recommendations through geolocation and soil data, and multilingual digital advisory tools. These tools sought to address the digital and literacy barriers that marginalized, elderly, and women farmers grapple with every day.

The voice-enabled and multilingual features significantly improved accessibility and system usability. They enabled wider uptake of advisory services, enhanced program discovery, and more informed decision-making. The result is a more inclusive digital transformation in agriculture, which demonstrates how human-centric AI can bring critical services to underserved farmer communities across Bihar.

The Gates Foundation commissioned this project.

SoBanHang is Vietnam’s leading B2B business app, which enables more than 500,000 merchants to manage sales, inventory, and digital finances on their smartphones. The app collectively processes approximately USD 2.5 billion in annual transactions. The institution acknowledged the lack of robust credit scoring models for its merchant base. It sought to develop a data-informed, predictive framework to assess credit risk and improve portfolio quality.

MSC collaborated with SoBanHang to create a sophisticated ML-driven credit scoring system customized for micromerchants. The project included rigorous model validation methods, such as, careful tuning, cross-validation, and real-world testing, to ensure predictive reliability. MSC also designed a merchant self-assessment tool to support financial inclusion and transparency. This tool enabled entrepreneurs to gauge their creditworthiness firsthand. Field pilots with merchants helped refine usability and effectiveness.

This work significantly enhanced SoBanHang’s risk assessment capabilities, enabled more confident lending decisions, and increased transparency for merchants. The introduction of the self-assessment tool also promoted greater financial awareness among users. This catalyzed a stronger and more inclusive digital lending ecosystem in Vietnam.

SoBanHang commissioned this project

Frontier Markets operates an extensive rural distribution network powered by women microentrepreneurs known as Sahelis. They deliver essential products and services to more than 4.3 million rural customers across 5,000+ villages, and operate through 35,000 digitized outlets. Despite this impressive reach, variability in agent performance and limited targeting of high-potential customers posed significant risks to impact and efficiency.

MSC partnered with Frontier Markets to address these interconnected challenges. We deployed a dual-strategy solution that blended AI-driven profiling with predictive customer segmentation. We applied machine learning methods to the performance data of Sahelis. This data included sales, engagement, and demographics to identify early indicators of success and integrate a profiling model into recruitment and retention workflows. We also developed customer segmentation tools based on unsupervised clustering and supervised classification. These tools enabled Frontier Markets to predict customer lifetime value from early behavior patterns and adapt outreach accordingly.

The results were remarkable. Frontier Markets focused on agents most likely to succeed and prioritized high-value customers through personalized engagement. This approach reinforced agent retention, improved service relevance, and optimized marketing and onboarding efforts. It amplified reach, promoted reliability, and reinforced operational sustainability and rural impact.

JPMorgan Chase commissioned the project.

Millions of blue-collar workers in India lack formal credit histories. This creates a challenge for financial institutions when they attempt to assess creditworthiness and offer suitable products. Entitled is a digital platform that offers financial services to more than 34 million blue-collar workers across India. It partnered with MSC to extract meaningful financial insights from unstructured mobile data and bridge this gap.

MSC applied natural language processing (NLP) and text mining techniques to analyze mobile SMS data, such as banking alerts and spending notifications. This allowed our experts to identify key financial behaviors, which included income patterns, savings activity, and repayment history. We used the insights to create structured customer profiles. The profiles enabled Entitled to design more relevant financial products for its users. MSC also developed a user-friendly dashboard to help Entitled scale this approach, monitor user behavior, and continuously refine service delivery.

Entitled used the project to serve low-income and thin-file customers better. It turned behavioral data into actionable financial intelligence and improved access to savings and credit solutions for a traditionally underserved segment. This approach offered a path to responsible financial inclusion for users without formal banking histories. It also aligned products with the actual financial capacity and habits of these users.

JPMorgan Chase commissioned the project.