[caption id="attachment_91792" align="alignleft" width="225"]
Urmila Chandrasekeram[/caption]
[caption id="attachment_91793" align="alignright" width="225"]
Kukaraj Tharmasegaram[/caption]
By Uwin Lugoda
Credit plays a major role in a country’s economy at the best of times, but in an economic crisis such as the one Sri Lanka is facing at the moment, credit can be of help to businesses and individuals who are on the verge of bankruptcy and losing jobs to survive and move towards better days. For this purpose, the Government and the Central Bank of Sri Lanka (CBSL) have been striving to boost liquidity in the banking sector and ensure that liquidity is transmitted in the form of loans to businesses and individuals impacted by Covid-19.
However, with the increased need for credit in the country and the constant requests by the Central Bank for banks to increase lending comes enhanced credit risk. Sri Lanka’s non-performing loans (NPLs) were already high prior to Covid, and indiscriminate lending could make NPLs shoot up even further.
This is why Algoredge, a Sri Lanka-based start-up, has now introduced “Mint Augmented Machine Learning” (MAML), a platform that aids local banks and non-banking financial institutions (NBFIs) in making better lending decisions.
Launched in early June, MAML was the brainchild of two electrical and computer systems engineers, Kukaraj Tharmasegaram and Urmila Chandrasekeram, who wanted to help Sri Lanka’s financial sector.
Speaking to The Sunday Morning Business, Chandrasekeram, who is in charge of the business aspect of the platform, stated that in Sri Lanka, banks as well as NBFIs are in competitive pressure to increase lending and drive customer acquisition. She explained that the downside to this is the staggering credit risk that could cripple an organisation’s overall performance.
“One of the major causes for this is rooted in the existing credit scoring method where the loan decisioning is made with only a handful of data about the borrower.”
Chandrasekeram stated that while this method works well when evaluating borrowers with long-term credit histories, for those without similar histories, the method is not able to differentiate between creditworthy and high-risk borrowers.
She stated that in 2019, they embarked on a journey to launch a “buy now, pay later” payment platform called Mintpay which allowed shoppers to split the total purchase cost into installments at the checkout. She went on to state that MAML was actually the technology they developed for Mintpay to assess the creditworthiness of the individual using their digital footprint.
“After developing MAML, we realised this technology will be used in the financial sector to improve credit decisioning.”
Tharmasegaram, who overlooks the technology aspect of the platform, stated that their current solution is crafted to assist financial institutions to identify better borrowers and reduce loan losses by using big data analytics and machine learning.
In this they follow a four-step process where the duo first gather data from the financial company’s database and label them in order to gather the characteristics and features of the dataset. This dataset is then analysed on multiple iterations before landing on an optimal machine learning algorithm that meets the organisation's target. Once the model is finalised, it is deployed either on cloud or on-premise. This algorithm will then continue to evolve with consumer trends and the economic climate to deliver an accurate assessment.
The financial sector, which is composed of banks as well as NBFIs, cater to different demographics. Tharmasegaram stated that their solution, MAML, can be tailored to serve the specific needs of the organisation, whether it is to reduce credit loss, identify better borrowers, or mitigate risk.
Speaking as to how they generate revenue through MAML, Chandrasekeram stated that they license their technology to financial institutions on a subscription basis.
MAML is also one of the fintech start-ups that received support from HatchX, Sri Lanka's first fintech accelerator programme carried out by Hatch in partnership with Lankan Angel Network (LAN). According to Chandrasekeram, HatchX helped them gain invaluable knowledge and a great network of experts in the financial services sector in Sri Lanka.
“The programme, through great mentors, teachers, and speakers, has helped us to strategise our start-up based on a rigorous analysis, together with metrics to monitor and quantify the results of the actions. Overall, I would say it is a well-rounded accelerator programme that paves the way to expand our network and create new opportunities.”
Finally, Chandrasekeram stated while MAML is currently geared towards helping mitigate credit risk and making credit decisions better and easier for banks and financial institutions, its second phase will work towards boosting financial inclusivity in Sri Lanka, which has been seen to be a prevailing problem not only in Sri Lanka but in all emerging markets.
Urmila Chandrasekeram[/caption]
[caption id="attachment_91793" align="alignright" width="225"] Kukaraj Tharmasegaram[/caption]
By Uwin Lugoda
Credit plays a major role in a country’s economy at the best of times, but in an economic crisis such as the one Sri Lanka is facing at the moment, credit can be of help to businesses and individuals who are on the verge of bankruptcy and losing jobs to survive and move towards better days. For this purpose, the Government and the Central Bank of Sri Lanka (CBSL) have been striving to boost liquidity in the banking sector and ensure that liquidity is transmitted in the form of loans to businesses and individuals impacted by Covid-19.
However, with the increased need for credit in the country and the constant requests by the Central Bank for banks to increase lending comes enhanced credit risk. Sri Lanka’s non-performing loans (NPLs) were already high prior to Covid, and indiscriminate lending could make NPLs shoot up even further.
This is why Algoredge, a Sri Lanka-based start-up, has now introduced “Mint Augmented Machine Learning” (MAML), a platform that aids local banks and non-banking financial institutions (NBFIs) in making better lending decisions.
Launched in early June, MAML was the brainchild of two electrical and computer systems engineers, Kukaraj Tharmasegaram and Urmila Chandrasekeram, who wanted to help Sri Lanka’s financial sector.
Speaking to The Sunday Morning Business, Chandrasekeram, who is in charge of the business aspect of the platform, stated that in Sri Lanka, banks as well as NBFIs are in competitive pressure to increase lending and drive customer acquisition. She explained that the downside to this is the staggering credit risk that could cripple an organisation’s overall performance.
“One of the major causes for this is rooted in the existing credit scoring method where the loan decisioning is made with only a handful of data about the borrower.”
Chandrasekeram stated that while this method works well when evaluating borrowers with long-term credit histories, for those without similar histories, the method is not able to differentiate between creditworthy and high-risk borrowers.
She stated that in 2019, they embarked on a journey to launch a “buy now, pay later” payment platform called Mintpay which allowed shoppers to split the total purchase cost into installments at the checkout. She went on to state that MAML was actually the technology they developed for Mintpay to assess the creditworthiness of the individual using their digital footprint.
“After developing MAML, we realised this technology will be used in the financial sector to improve credit decisioning.”
Tharmasegaram, who overlooks the technology aspect of the platform, stated that their current solution is crafted to assist financial institutions to identify better borrowers and reduce loan losses by using big data analytics and machine learning.
In this they follow a four-step process where the duo first gather data from the financial company’s database and label them in order to gather the characteristics and features of the dataset. This dataset is then analysed on multiple iterations before landing on an optimal machine learning algorithm that meets the organisation's target. Once the model is finalised, it is deployed either on cloud or on-premise. This algorithm will then continue to evolve with consumer trends and the economic climate to deliver an accurate assessment.
The financial sector, which is composed of banks as well as NBFIs, cater to different demographics. Tharmasegaram stated that their solution, MAML, can be tailored to serve the specific needs of the organisation, whether it is to reduce credit loss, identify better borrowers, or mitigate risk.
Speaking as to how they generate revenue through MAML, Chandrasekeram stated that they license their technology to financial institutions on a subscription basis.
MAML is also one of the fintech start-ups that received support from HatchX, Sri Lanka's first fintech accelerator programme carried out by Hatch in partnership with Lankan Angel Network (LAN). According to Chandrasekeram, HatchX helped them gain invaluable knowledge and a great network of experts in the financial services sector in Sri Lanka.
“The programme, through great mentors, teachers, and speakers, has helped us to strategise our start-up based on a rigorous analysis, together with metrics to monitor and quantify the results of the actions. Overall, I would say it is a well-rounded accelerator programme that paves the way to expand our network and create new opportunities.”
Finally, Chandrasekeram stated while MAML is currently geared towards helping mitigate credit risk and making credit decisions better and easier for banks and financial institutions, its second phase will work towards boosting financial inclusivity in Sri Lanka, which has been seen to be a prevailing problem not only in Sri Lanka but in all emerging markets.