The technology behind the MSME Fintech Startup ARTH
According to media reports, the micro, small and medium-sized enterprises (MSMEs) sector is currently facing a credit deficit of $ 380 billion. Mumbai-based ARTH wants to use technology to bridge this gap.
Women’s fintech start-up MPME was founded in 2018 by Shweta Aprameya. Established under parent company Arthimpact Digital Loans, the startup focuses on women-led businesses with little or no online adoption. It started out by offering credit services and quickly added built-in insurance to its product portfolio. So far, it has granted more than five lakhs of loans, affecting more than two lakhs of micro-entrepreneurs.
The startup assesses clients using reliable and verified data, occupation, digital records and social factors and does not rely on standard documents such as credit bureau records or tax returns on Income. The fintech startup operates in micro-markets and has served 3.5 lakh micro-entrepreneurs who do not have bank statements or GST statements on 18,400 PINs in India.
Recently, product manager Raman taneja engaged in a conversation with Analytics India Magazine to discuss how ARTH uses AI and ML create digital personas for clients that ultimately help design fit financial products, rather than taking a one-size-fits-all approach.
Raman holds an MBA from IBS Hyderabad and previously worked as CTO at Dvara Solutions and as Head of Digital Financial Inclusion, Consumer Banking, Strategic Partnerships and Distribution Channels at Jio Payments Bank.
Edited excerpts from the conversation:
OBJECTIVE: How painful is ARTH trying to respond? How is it?
Raman Taneja: ARTH targets low credit micro-MSMEs (NTC) entrepreneurs and micro-MSMEs. These include micro-businesses in all categories including grocery stores, clothing, restaurants and others. While these entrepreneurs know how to manage their money, they lack access to formal credit and other relevant financial services such as insurance and digital payments.
ARTH’s Theory of Change is based on the premise that if fair and flexible financial access is granted to microentrepreneurs, it can generate viable returns on investment.
Since the adoption of the technology is still limited among the target group it serves, we have adopted a hybrid touch and technology model. The team also helps new credit customers understand the details of the credit and payment mechanism to ensure there are no surprises for the customer down the road. In turn, this allows us to better understand a customer’s needs and design products that are right for them.
For example, during the Covid-19 lockdown, many customers needed capital to restart their businesses. We have offered âCOVID Rahat Loans,â interest-free financial loans to help severely affected customers. Rahat loans targeted women nano-entrepreneurs whose income is less than Rs 3 lakh per year.
OBJECTIVE: What are your main roles and responsibilities as CPO at ARTH?
Raman Taneja: My responsibilities are to develop technology-driven products and partnerships to have an easy-to-use interface for both digital and non-digital customer segments. The product technology stack makes extensive use of open source technology and has adopted innovations from around the world. Additionally, we use public and private APIs to collect customer data and facilitate navigation throughout the onboarding experience.
The idea is to create a technological stack of products to manage scale. The product roadmap includes creating a seamless credit and insurance experience across partners and channels. Although our focus so far is on credit, we will soon be expanding to other financial products for high-end micro-MSMEs through our digital platform.
OBJECTIVE: What are your main products and services?
Raman Taneja: At present, the focus has been on providing credit products to MSME micro-entrepreneurs. We have co-developed innovative insurance products with one of the leading digital insurance companies, focusing on the different needs of micro-MSMEs. Through insurance products, we provide life and health insurance as well as income protection to micro-MSMEs.
We also plan to launch new products for the upscaling of micro-MSMEs via digital platforms. The product roadmap includes creating a seamless credit and insurance experience across partners and channels. The digital platform will offer links to upstream and downstream supply chains to make credit available on demand.
OBJECTIVE: Explain the technology stack.
Raman Taneja: We used a combination of open-tech platforms like Angular, Ruby on Rails and Postgres to develop our technology backend.
Open source platforms allow us to adopt and implement some of the world’s best-developed technologies at a reasonable cost of ownership. This includes mobile frontend frameworks and backend technology. The technical stack is API and microservices framework compliant and can be integrated with any number of APIs within days.
Raman Taneja: We use:
- Open source tools for cloud development and deployment
- The new framework under development is also cloud native
- Tools provided by Amazon Web Services to manage day-to-day operations
- Optical character recognition (OCR) and computer vision tools to read KYC documents.
OBJECTIVE: Explain the ML model you use to detect fraud.
Raman Taneja: We use reliable open source data available publicly as well as alternative data collected from users to train our AI and ML models. All of the customer lifecycle data is fed into the ML model to train our engine to identify potential risks and new opportunities.
The ML model trains on customer onboarding, transaction, and public API data to create predictive behavior, which helps to better identify risks.
AIM: How do you leverage AI and ML technologies?
Raman Taneja: To take out a customer for credit, our proprietary ML model determines eligibility using alternative social, economic, business and behavioral data. To improve and continue deep learning of the customer base, the ML model evolves a predictive risk score of a customer and identifies patterns for payers and non-payers.
Since ARTH largely works with micro-entrepreneurs new to credit and micro-entrepreneurs, we place great importance on educating and educating clients. Once the credit is approved, our representatives pass their know-how on to each customer and explain the details of the interest payment schedule and other rules such as late fees. Comments from each of these sessions are also fed into the ML model for future use.
We use computer vision to ensure that documents completed by a client for KYC are valid and will be acceptable as directed by RBI. This helps our systems detect any anomalies in the submission in real time.
OBJECTIVE: Who do you see yourself competing with? How do you differentiate yourself from your competition?
Raman Taneja: We have remained focused on meeting the credit needs of micro-MSMEs. Most of these companies do not have financial information in standard formats, our AI / ML engine captures data across social, economic and business parameters to assess creditworthiness. Our credit size is also extremely flexible to meet a client’s needs at different times.
We reach out to our potential customers through technological and tactile models and work with network operators at national and hyperlocal levels.
OBJECTIVE: What is the success rate of your model?
Raman Taneja: The model has shown great results with smaller clients, demonstrating the goal of having a large-scale impact on the micro-MSME segment. The favorable experience in accessing credit translates into a large loyal base of over 60 percent of loyal customers. ARTH’s credit risk model has so far shown one of the best results in the industry.
OBJECTIVE: How many loans have been disbursed by ARTH so far? What is the average ticket size?
Raman Taneja: ARTH has served 350,000 unique clients, and 55% of them are new to credit. The average ticket size for a loan is around Rs 50,000.
OBJECTIVE: What are ARTH’s future projects?
Raman Taneja: Financial inclusion in India has made great collective strides. From a largely unbanked country, the various initiatives supported by regulators and the government have led to the opening of nearly a billion bank accounts in India and access to customers through a large number of bank agents. However, access to credit remains limited, especially for micro-enterprises and enterprises. These clients do not have a formal record of their business transactions and establishments, leading to typical problems of information asymmetry and exclusion.
Although our focus is on credit, we will soon be expanding to other financial services for high-end micro-MSMEs via a digital platform. ARTH will create an ecosystem of partners to provide on-demand credit to all stakeholders.
Subscribe to our newsletter
Receive the latest updates and relevant offers by sharing your email.