A walk-through the positive steps taken by a leading financial services provider to streamline and improve their debt recovery processes
When it comes to debt recovery from clients, the traditional methods of collection may be redundant, especially in today’s technology-driven age.
Leveraging AI and ML Technology is a great way to achieve stellar results by employing strategic tweaks in the debt recovery processes.
Read on to know how CreditNirvana assisted with the seamless integration of technology to achieve best in class outcomes.
A very large Mortgage financial services provider, in the India with total assets of US$ 7.2 Billion
The client was facing challenges in collections efficiency, effectiveness, customer satisfaction and increasing collection costs.
Most processes were manually intensive resulting in lengthy turnaround times. The collection prioritization process was primarily intuitive, demographic and bureau score driven, resulting in wrong accounts being chased for collections.
There has been a high level of dependency on costly channels, such as phone calls to communicate with customers. There was also no feedback mechanism to improve the customer contact strategy.
The client chose CreditNirvana, with our deep focus in platform driven DeepTech, digital-led execution and deep domain expertise. We conducted studies on their collections operations that included historical data evaluation and stakeholder interviews, and then deployed CreditNirvana in the private cloud infrastructure of the organization.
The debt collection related data were spread across more than 7 different IT systems and database scopes. Using the CreditNirvana API and database connectors, we integrated all these data into CreditNirvana Big Data debt collection platform. While the individual Credit Departments of each loan portfolio has their independent view of the data, at an organizational level, a consolidated data of all
Customers was created which also gave a 360˚ view of all the assets of each customer.
The bank was segmenting its customers into collection risk levels using limited, internal collections structured data; however, the outputs were not dynamic and real-time. The existing analytics were not using NLP/NLG techniques for customer analytics. To improve segmentation and personalization, we revisited the predictive models by leveraging CreditNirvana machine learning models and bringing in unstructured data (call center data, mobile application data and Digital follow up data) and external data sets such as industry indices data and bureau data.
By leveraging the power of CreditNirvana ML engine and by tweaking the existing models using additional data sets, bringing in 300 plus featured variables of CreditNirvana ML engine and leveraging Precision NLP/NLG techniques, very robust ML models were deployed resulting in providing real-time and dynamic predictive and prescriptive outputs. These outputs have strengthened the collection segmentation, prioritisation and personalization of collection follow ups.
While the predictive outputs such a Payment scores, Payment bands, expected payments, and expected payment dates are helpful in improved segmentation and prioritization, the Prescriptive outputs such as Customers’ preferred channel, preferred time of contact and personalized content for follow-up are helpful in automated Digital first collection follow-up execution.
Robotic process automation was implemented across all collection processes, and in all delinquency levels. The prescriptive outputs from the Precision ML Engine were automatically executed through various digital channels including integrating with payment gateway engines.
Multi-lingual AI Conversational engine was deployed in messenger channels and integrated with digital payment engines.
The seamless process execution of Robo collection process is being managed by CreditNirvana collection execution team and with this, we are fulfilling the end to collection processes for our clients.
Collection management systems have been deployed to provide easy management of all the collection follow-up, payment collection and settlement.
Various workflows for collection follow-up and payment collection manual collection process and legal management have also been deployed.
The allocation to collection agencies/agents were more of intuition-driven and less of advanced analytics-driven. This in turn, resulted in chasing wrong accounts as a priority and hence brought along several inefficiencies.
CreditNirvana’s personalized output helped in accurate prioritization of accounts to be followed up and thus improved the efficiency.
The above, along with reaching out to customers on their preferred date /time resulted in enhanced positive responses from the customers and reduced the number of times customers needed to be reached out to.
The client was able to manually monitor less than 5% of agent calls. The CreditNirvana NLP/NLG engine is programmed to analyze 100% of calls and the insights are being used to enhance agent performance. These outputs are also helping in customized agent training to ensure compliances and also to improve customer experiences.
Further, CreditNirvana Robotic process automation aggregates customer information from multiple sources in real-time and shown in one informative dashboard. This is helping agents to access relevant information seamlessly, on their handheld instruments, and this in turn helps in servicing the customers more efficiently.
Robotic process automation process helping the collection departments to streamline operations and improve efficiency. By leveraging AI Conversational engine and bots, appropriate actions for all customer interactions are taken dynamically and in real-time using straight through processing.
For example, if a customer confirms that he would be receiving payments from a third party on a particular date, bots are engaged to monitor this and a digital payment engine shall be sent to him/her through his/her preferred channel on the same day at the prescribed time with a recorded message of his earlier promise. If it turns out that the customer had misled the bank, appropriate actions shall be executed automatically. In this case, repayment can be collected without any human intervention at all, making this scenario an ideal one for straight through processing.
Traditionally, the client was relying on agent calls as the primary method of debt collection.
CreditNirvana Robo Execution engine provides the AI conversational interactions with customers through messengers and ‘robo’ calls for reaching out to customers through their preferred channels, at their preferred date & time, which has made customer experiences very pleasant.
Along with this, the integrated Digital payment engine has made the follow-up and payment process intelligent and very user-friendly. Real- time personalized settlement offering and closure through Mobile apps has also made the customer experience extremely satisfying.
CreditNirvana Robo collection process helps in lowering the in-person agent handled accounts, reduction in average handling time, increased collection rates and improved customer experiences.
The calling-agent driven collection process was causing errors and compliance breaches.
CreditNirvana follow-up personalisation process and AI conversational engine, made through pre- approved templates of communication processes, is eliminating the human emotions and errors in the customer interaction process.
This process has substantially strengthened the compliance processes.
After 12 months of deployment, we have jointly conducted an ‘impact study’ with the bank and following results were noted:
We help clients improving their collection management process, leveraging the power of data and digital technologies. We work closely together with our clients, managing debt collection in evolving business eco-systems and fast-changing economic circumstances, enabling them to fully focus on their core activities. Whether you are in Consumer Finance, Commercial Finance, Credit Union, Community banks, Credit Cards, Healthcare, Utilities, or Telco, CreditNirvana has tailor made ML driven debt collection platform solutions & services to help you!