Financial data analytics engine associated with a customer relationship management system
Methods, systems, and computer storage media for providing financial data analytics recommendations using a data analytics engine in a customer relationship management system. The recommendations can be a lead that is information associated with a model-generated suggested consumer solution, an alert of increased risk of attrition, or an alert of increased risk of default. The data analytics engine is configured to generate target variables associated with financial products or the customer relationship and utilize modeling techniques and apply rules to generate recommendations. Operationally, the recommendations are generated based on a data analytics model. Generating the recommendations is based on feature variables that are generated based on aggregation and transformation of customer data and utilizing machine learning models to detect patterns in the customer data using the feature variables. The recommendations can be presented via a financial data analytics interface along with insights that provide plain text explanations of the recommendations.
1 . A computerized system comprising:
one or more computer processors; and
computer memory storing computer-useable instructions that, when used by the one or more computer processors, cause the one or more computer processors to perform operations comprising:
generating, by the one or more processors, training data based on first input data comprising customer data, transaction data, and product data, wherein generating the training data comprises:
aggregating the first input data to generate a plurality of data chunks, wherein each data chunk represents an aggregate of a respective portion of the first input data over a respective time period and for a respective customer,
defining a backward window time period and a forward window time period,
generating a plurality of feature variables based on the data chunks, the plurality of feature variables representing features of the first input data corresponding to the backward window time period, and
generating a plurality of target variables based on the first input data, the plurality of feature variables representing events occurring during the forward window time period;
training, by the one or more processors using the training data, a predictive machine learning model to perform operations including generating a first financial data analytics recommendation associated with a customer risk of attribution and a second financial data analytics recommendation associated with insights associated with one or more customers or products,
wherein training the predictive machine learning model comprises:
aggregating at least some of the feature variables with respect to the backward window time period, and
determining one or more patterns between the feature variables and the target variables;
accessing, by the one or more processors, second input data of a customer;
analyzing the second input data using the predictive machine learning model;
rendering, on a display device, a graphical user interface for presentation to the user, wherein rendering the graphical user interface comprises:
determining a set of financial data analytics recommendations including a first financial data analytics recommendation associated with the customer,
grouping, in the graphical user interface, visual representations of the set of financial data analytics recommendations according to a quality of each of the financial data analytics recommendations, wherein the quality of each of the financial data analytics recommendations is determined using the predictive machine learning model,
filtering, from the graphical user interface, at least some of the visual representation of the financial data analytics recommendations based on one or more filter rules, and
causing rendering, in the graphical user interface, of at least some of the grouped and filtered financial data analytics recommendations including the first financial data analytics recommendation, along with one or more visual representations of a plurality of insights that provide human-readable explanations for the first financial data analytics recommendation.
2 . The system of claim 1 , wherein training the predictive machine learning model comprises aggregating the training data based on selected data aggregation levels and product aggregation levels.
3 . The system of claim 1 , wherein the predictive machine learning model is trained based on a tree-based approach that identifies customer segments as subgroups having different target rates based on feature variable values of customers.
4 . The system of claim 1 , wherein the first financial data analytics recommendation is generated based in part on assigning the customer to a subgroup based on feature variable values of the customer and a predicted likelihood that the customer corresponds to a target rate of the subgroup.
5 . The system of claim 1 , the operations further comprising assigning a financial product lead information insight that explains why the customer is likely to purchase a corresponding product during a specified time period.
6 . The system of claim 1 , the operations further comprising assigning a financial product lead information insight that explain why the customer is to churn within a specified time period.
7 . The system of claim 1 , the operations further comprising calculating an expected monetary impact of the first financial data analytics recommendation.
8 . The system of claim 1 , wherein the first financial data analytics recommendation comprises financial product lead information including a specific product and a specific point in time that is included in an insight from the plurality of insights.
9 . The system of claim 1 , wherein the graphical user interface comprises financial data analytics interface elements associated with solution interface data.
10 . The system of claim 1 , wherein the graphical user interface comprises financial data analytics interface elements associated with retention interface data.
11 . The system of claim 1 , wherein the predictive machine learning model is fitted to the feature variables and the target variables of a plurality of products to identify a best fit model for the plurality of products or an alert for increased risk of attrition of the customer.
12 . The system of claim 1 , the operations further comprising:
causing generation of the first financial data analytics recommendation and the second financial data analytics recommendation based on:
detecting, using the predictive machine learning model, patterns in input data associated with the feature variables;
based on the detected patterns, generating values for target variables that are based on values of the feature variables;
based on generating the values for the target variables, generating a global set of financial data analytics recommendations having lead information associated with the feature variables and the target variables applying business rules based on a scoring model to filter the global set of financial data analytics recommendations;
applying overlay rules to suppress or change the global set of financial data analytics recommendations; and
communicating, for presentation on the graphical user interface, the first financial data analytics recommendation and the second financial data analytics recommendation.
13 . A computer-implemented method, the method comprising:
generating, by one or more processors, training data based on first input data comprising customer data, transaction data, and product data, wherein generating the training data comprises:
aggregating the first input data to generate a plurality of data chunks, wherein each data chunk represents an aggregate of a respective portion of the first input data over a respective time period and for a respective customer,
defining a backward window time period and a forward window time period,
generating a plurality of feature variables based on the data chunks, the plurality of feature variables representing features of the first input data corresponding to the backward window time period, and
generating a plurality of target variables based on the first input data, the plurality of feature variables representing events occurring during the forward window time period;
training, by the one or more processors using the training data, a predictive machine learning model to perform operations including generating a first financial data analytics recommendation associated with a customer risk of attribution and a second financial data analytics recommendation associated with insights associated with one or more customers or products,
wherein training the predictive machine learning model comprises:
aggregating at least some of the feature variables with respect to the backward window time period, and
determining one or more patterns between the feature variables and the target variables;
accessing, by the one or more processors, second input data of a customer;
analyzing the second input data using the predict machine learning model;
rendering, on a display device, a graphical user interface for presentation to the user using a web-service, wherein rendering the graphical user interface comprises:
determining a set of financial data analytics recommendations including a first financial data analytics recommendation associated with the customer,
grouping, in the graphical user interface, visual representations of the set of financial data analytics recommendations according to a quality of each of the financial data analytics recommendations, wherein the quality of each of the financial data analytics recommendations is determined using the predictive machine learning model,
filtering, from the graphical user interface, at least some of the visual representation of the financial data analytics recommendations based on one or more filter rules, and
causing rendering, in the graphical user interface, of at least some of the grouped and filtered financial data analytics recommendations including the first financial data analytics recommendation, along with one or more visual representations of a plurality of insights that provide human-readable explanations for the first financial data analytics recommendation.
14 . The method of claim 13 , wherein the financial data analytics recommendation comprises financial product lead information including a specific product and a specific point in time that is included in an insight from the plurality of insights.