The Challenge

The client was in the business of collections with reference to Consumer delinquencies. The collections process was initiated over a call with the goal to receive a Promise to Pay with reference to a specific amount. The key here is to ask for the right amount to get the party to agree to pay. This step wise process which involves getting the amount from the party, split up into different sums, eventually aims towards total loan amount recovery.

The challenge here was that the executives had to ask for the correct amount from the customer which would maximise the probability of the party agreeing to it and giving a promise to pay. Asking for an amount which did not drive the party to agreeing for making the payment defeated the whole purpose of the exercise and eventually lead to unsuccessful calls, wasted resources and strayed further away from the eventual goal of loan recovery.

The Solution

PredictSense used AI to come up with the perfect solution for this challenge. The focus of the solution was on providing the executives with the input of the right amount to ask for getting the Promise to Pay.

The system worked on building a model which was based on Machine Learning. The client provided data points which included the debtors profile, income and demographic details, previous records with reference to collections like amount promised and honoured, outstanding amount, failed attempts and more. Based on this data the system generated a model. The output of the model was the right amount to pay to be asked from a particular debtor to maximise the probability of getting a promise to pay. The model considered various factors and came up with the right solution. The ultimate result of every prediction was taken as an input and used by the model to optimise its system. Thus the model was dramatically improving its accuracy as the feedback loops were established.

This AI driven approach to the collections process had a tremendous effect on the loan recovery rate. The executives were making calls using a very efficient method and thus they were high impact calls. The results were seen in terms of an increase in promise to pay which were honoured and led to faster loan recovery. The executives were now even more motivated and driven to achieve their objectives using the intelligence of PredictSense. Thus PredictSense was a game changer for the client’s collections business.


  • 75% faster
    collections rate

  • 45% increase
    in call success

  • 60% increase
    in productivity


1. Faster recovery of loan amounts from debtors because of efficient calling

2. Reduced costs as executives successfully closed on calls faster with the help of PredictSense

3. Increased work satisfaction leading to better performance. Overall increase in efficiency reduced business costs

Want us to help you transform your business?

Register to start your transformation journey now

Case studies

  • Collection cost optimization with machine learning

    The client was in the business of collections with reference to Consumer delinquencies.

  • Intelligent collection amount recommendation system

    The client was in the business of collections with reference to Consumer delinquencies.

  • AI based tenant scoring solution

    US based tenant screening company was using a legacy software to assess the credibility of applicants seeking rental properties.