Friday, December 9, 2022
HomeManagementCrediwatch Assesses SME Creditworthiness Through Al & Machine Learning

Crediwatch Assesses SME Creditworthiness Through Al & Machine Learning

The fund-starved SME sector can benefit immensely if its financial and non-financial data points are assessed judiciously.

The need for Artificial Intelligence and machine learning for assessing creditworthiness struck Meghna Suryakumar and Sandeep Anandampillai while they were providing legal and IT services to legal and management consultancy firms. Both had deep experience with technology in their respective domains – Meghna as a transborder M&A legal consultant and Sandeep as an ace IT platform builder. Convinced with their notion, in 2016 they set up Crediwatch to offer data insights as a service. Through their proprietary technology, they would use millions of financial and non-financial data points to build trust scores for enterprises. Through their tech platform they would help their clients make informed decisions about vendors, buyers and borrowers in matters of credit.

In particular, the duo feels that through Crediwatch they can bring succour to the fund-starved SME sector. “We found gaps in the market which can only be addressed if the credit providers have valuable data-points of these small and medium businesses. We realised that focusing on risk and business insights by using a proprietary AI-based predictive engine is the way forward for the financial services industry” says Meghna, founder & CEO of Crediwatch, in an interview with Sourcing Hardware.

How has the journey been so far in terms of scale of business and financial milestones? What is your vision for the future?
Crediwatch has grown significantly by almost doubling in size with respect to employees as compared to the last calendar year. We now have 63 employees. We have partnered with very respected financial institutions, to provide them credit intelligence of more than 80 million (8 crores) insights about 50,000 businesses that have a collective asset base of $7 billion. Several leading public and private sector banks along with large and medium NBFCs are currently availing one or many of our solutions.

We have gone live with our Early Warning Signals (EWS) platform at Karur Vysya Bank and Aditya Birla Finance, two large marquee banking or lending institutions in the country. These were first-of-a-kind implementations, where the system captures early signs of distress by gleaning into both private account conduct data as well as public data footprints. Backed by our recent wins and a robust pipeline, Crediwatch is poised to double its revenues from last year. In the future, we will focus on enhancing our offerings to the SME segment, where we see tremendous potential. This is one segment that can significantly impact the GDP of the country – and our focus is to be a primary driver.

It is said that the SME sector is suffering due to massive funding gap. What are the reasons for the situation? How is technology helping to direct finances towards the deserving, smoothly and efficiently?
I think that the funding progress has been tardy so far for MSMEs and SMEs, even before COVID-19. The credit extended by banks to the MSME sector grew only 2.6% YoY as of February 2020. In the current situation, the SMEs and merchants will be further squeezed for funds in the coming months. All the lenders must evaluate loan applications at a much faster speed to increase the disbursement, of course with prudent measures in place.

Typically, the credit evaluation process for a secured loan at a public sector bank takes about 4-6 weeks. This can only come down with the use of technologies such as AI and machine learning, which bring insights into the credit underwriting process. Credit-risk analysis is another area where the deployment of machine learning will allow lenders to assess the true creditworthiness of borrowers. This technology can identify intricate, nonlinear patterns within large volumes of data, which helps differentiate legitimate borrowers from fraudsters. The most crucial advantage that machine learning offers is that it can learn continually. The more data an ML-based system is fed, the more accurate its credit risk predictions will be

How can Crediwatch help enterprises in the building products and home improvement (BPHI) industry improve credit assessment of their channel partners and institutional buyers?
When one uses the credit reports from Crediwatch, it becomes feasible, both in terms of cost and scale, for an enterprise to perform a just credit assessment of any of its channel partners or institutional buyers. By presenting a comprehensive and exhaustive view of public data on an entity, Crediwatch helps in giving its clients a superior edge at reducing exposure to risky engagements and assists them in choosing partners wisely. This analysis can be further enhanced if one automates the analysis of the partner’s GST returns and bank statements after using their consent. Considering the BPHI sector, where the number and types of suppliers are quite high, the use case of Crediwatch’s credit assessment solutions becomes even more meaningful. The segmentation of prospective channel partners and institutional buyers can be turned into an automated exercise with ease, leading to advantages of scale.

Also Read: How Escrowffrr is Putting Trust Back into Contingency Payments

What impact can scientific credit and risk assessment have on the growth and wellbeing of an enterprise? Note that the BPHI industry runs on long credit cycles, and channel finance is an essential aspect of the business.
In the present time of volatility, I feel that credit and risk assessment cannot be taken as a normative exercise. Using the flavours of AI and machine learning, Crediwatch has made the entire process seamless with minimal human intervention. This gives an enterprise a crisp and objective view of the financial wellbeing of any entity across domains.

In the context of the BPHI industry and its engagement with channel partners in long credit cycles, it becomes crucial for them to follow a proactive rather than reactive approach when it comes to credit and risk assessment. A continuous real-time credit monitoring tool along with assessment reports should prove beneficial to manage risk during the engagement.

Can you share case studies or testimonials to show how SMEs have benefited from your services?
While the majority of our current clientele are financial institutions, who use our solutions to lend to SMEs, one of our earliest customers is a medium-sized distributor of electrical products in Mumbai. When the owner of the business was facing challenges in managing its credit cycle, they subscribed to our platform to regularly assess all debtors and creditors to allow the right number of credit days on their invoices. This helped them bring down their credit cycles drastically and manage working capital efficiently. The customer continues to engage with us for more advanced analytics now, involving GST return analysis of its suppliers.

What are the ways in which you assess creditworthiness, considering that financial data of corporates and even SMEs is usually not available?
Crediwatch carries financial data for 1.8 million public limited, private limited and LLPs registered on the MCA. Besides, we gather c.2,500 data points from financial, non-financial and alternate datasets on an entity, as it leaves its digital footprint during the normal course of business. Several SMEs are also proprietorships and partnerships, for whom non-financial data related to the owners are readily available, such as legal cases, GST information, and news media. Several lenders also have access to financial information directly from the SME – which can be analysed using the Crediwatch engine. In our experience, creating a credit profile of any business in the country has not been a challenge to date, as the government moves towards digitisation of several data sources.

Also Read: COVID-19 Vs MSME: Financial Market Expert and Prest Loans CEO Ashok Mittal Offers Advisory for Businesses

What are the non-financial data and alternate data sets/digital footprints that assist in shaping a credit profile?
Let me cite some examples of non-financial and alternate data sets: legal cases against the business as well as the promoter/directors, GST filing information with and without consent, EPFO filings by the business highlighting the number of employees and payments, news sentiment from adverse media, and director involvements in other companies.

How accurate are your assessments? Do financial institutions also rely on your reports?
You see, human bias is a significant influence on the decision-making process that has previously resulted in loan frauds and NPAs. In our experience, AI has streamlined the credit underwriting process with little-to-no human intervention. A predictive analytics-driven vertical approach has enabled lenders to analyse quantitative and qualitative risk factors to create a comprehensive borrower profile for assessment. Our EWS platform can predict distress 12-18 months in advance, and this is being used by several leading banks and NBFCs in the country to monitor their credit portfolios. Our credit reports are used by leading public and private sector banks to assess the creditworthiness of borrowers at the time of underwriting. 

How are your services or technology different from those of other players in the fintech sector?
Crediwatch picks up data from 25,000 different points, which is the highest in the industry from already existing information in the regulatory framework, whereas a bank only looks up 200 data points while considering the decision for a loan. Thus, this scale is what sets Crediwatch apart from the crowd. The intelligence can further be deployed for risk management and analysis, due diligence and in future can be used for providing ‘Trust Score’ for SMEs who are new to credit.

When will you launch Trust Score? What size of SMEs would be able to use the service? How will Trust Score work?
Trust Score is in the product pipeline for H2 2020 and can be used by businesses of all sizes. Large corporates typically will use it for onboarding vendors/suppliers and performing due diligence. Mid-sized companies and SMEs will use it to decide credit cycles for their suppliers and manage their working capital efficiently. The low cost and scalable commercial model will allow firms of all sizes to subscribe to Trust Score.

Using the plethora of financial, non-financial and alternate datasets of a business, Crediwatch aims to build a single score – similar to a Credit Bureau score for an individual – which will provide an assessment of the creditworthiness of the entity. The score can also be monitored to assess the change in the risk profile of the business. I will be able to share more details after the product is launched.

How important is credit and risk assessment in the present scenario of uncertainty, and as an ease-of-doing-business parameter?
Our country has more than 50 million (5 crores) SMEs who have been facing a liquidity crunch even before the COVID-19 pandemic. Out of these, only 15% have access to formal credit due to the trust deficit that exists, and their lack of collateral. We believe the situation will worsen in the present scenario. While the government may relax certain ease-of-doing-business guidelines, we do not expect lenders to start funding a business on the back of fears of higher NPAs. The credit assessment processes would need to accommodate short-term and long-term changes that are expected in the business models of various industries, even as the situations improve.

RELATED ARTICLES

Most Popular