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 reach the right party and request payment. This is a high cost process because it involves time and resources. It is crucial to reach the right individual on the call. Speaking to any other person will not be sufficient because it would mean more follow-up calls and increased overheads. The entire process needs to be driven by efficiency to meet the objective of closing on collections at minimum cost.
This would mean minimum number of attempts and maximum impact. It is crucial to speak to the right person in order to achieve this. Right Party Contact (RPC) in minimum attempts was the key. The client was facing the challenge and it was important to solve this problem as this translated to costs and had an adverse effect on business profits.
PredictSense provided a great solution to this problem. While the focus of the solution was optimising the collections processing, this solution is applicable to improve efficiency during tele calling for other objectives as well.
The solution was based on a model generated by PredictSense. The client was sitting on a huge amount of client calls data. This included the details about the time of call, whether the right party answered the call, call details and more. Based on this data the system could predict the probability of success of establishing contact with the right party along with the preferred time to increase this probability. The executives who were calling were guided by PredictSense in terms of telling them the best time to speak to the right person. The model built using AI improved itself continuously, based on the inputs from latest call details and new records. Thus, it was never relying on older patterns and kept itself updated as new patterns emerged. The model could train and retrain itself thus always works towards process optimization.
This dynamic learning model achieved increased levels of accuracy as it was fed with more data. The result was a phenomenal increase in efficiency as agents could now reach the right party with minimum effort. They worked smarter and their productivity and morale received a tremendous boost. PredictSense empowered collections agents to achieve the business goal of optimising collections processes. Next gen collection processes will work on data analysis and predictive models. And with PredictSense the opportunities to optimize collection processes are endless.
1. Efficient and result oriented calling system was established
2. Executives were able to work efficiently and this led to cost reduction and had a positive impact on profits
3. Reaching the right party with the help of AI, led to faster collections processing thus directly making a positive impact on business goal
Paython and Jupyter notebbok based interface Focus on data science and building quality models Deploy and monitor models in production effortlessly
Paython and Jupyter notebbok based interface Focus on data science and building quality models Deploy and monitor models in production effortlessly
Paython and Jupyter notebbok based interface Focus on data science and building quality models Deploy and monitor models in production effortlessly