US 11,854,017 B2
Predictive modeling and analytics integration platform
Bharat Goyal, San Jose, CA (US); and Pavan Korada, San Mateo, CA (US)
Assigned to Zeta Global Corp., New York, NY (US)
Filed by Zeta Global Corp., New York, NY (US)
Filed on Feb. 8, 2021, as Appl. No. 17/170,848.
Application 17/170,848 is a continuation of application No. 15/147,615, filed on May 5, 2016, granted, now 10,997,604.
Claims priority of provisional application 62/157,342, filed on May 5, 2015.
Prior Publication US 2021/0182865 A1, Jun. 17, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 30/00 (2023.01); G06N 5/04 (2023.01)
CPC G06Q 30/00 (2013.01) [G06N 5/04 (2013.01)] 16 Claims
OG exemplary drawing
 
1. A computer implemented method, comprising:
selecting a plurality of dynamic models for evaluating a scoring request, wherein the dynamic models are stored on a scoring database;
deploying each of the dynamic models to one of a plurality of evaluators;
synchronizing the dynamic models, the synchronizing including at least determining a same version of the dynamic models is deployed and available to all evaluators;
receiving, at a scoring node, a scoring request for a score of a lead from at least one requester;
separating the scoring request into a plurality of scoring requests, wherein each of the scoring requests is assigned to one of the selected dynamic models wherein the request is separated by model and aggregate model results and sent to an evaluator queue;
combining results from each of the dynamic models;
evaluating the combined results wherein each of the evaluators sends a model evaluation response to the response queue; and
providing a response to the scoring request based on the evaluation of the combined results.