System and method for matching entities using machine learning

1 . A method for matching computer equipment or software entities using artificial intelligence, the method comprising:

determining, from a set of features describing entities, a subset of features which when input to a machine learning module allow the machine learning module to match a pair of entities based on the ability of the second entity in the pair to successfully take over responsibilities handled by the first entity in the pair, wherein the pair of entities have disjoint features, and wherein the determining the subset of features comprises using a decision tree model, the tree comprising: decisions based on whether an entity in a pair has a feature above a threshold;

periodically training a machine learning module using the determined features, wherein the machine learning module comprises a random forest model, and wherein training the machine learning module using the determined features comprises providing the machine learning module with pairs of entities and for each pair an amount of improvement occurring when the second entity in the pair took over the responsibilities of the first entity in the pair;

collecting, using one or more microservices, information regarding a set of entities;

for a first entity described by the determined set of features, using the machine learning module to determine a list of entities matching the first entity, each entity in the list of entities associated with a matching score, wherein the determining of the list of entities matching the first entity comprises removing a feature from the input to the machine learning module and determining the effect on accuracy of the removing, and wherein the first entity described by the determined set of features and each entity in the list of entities each comprises computer equipment or software to be replaced; and

automatically activating an entity in the list of entities.

2 . The method of claim 1 , wherein the matching score for an entity reflects the ability of the entity to successfully take over responsibilities handled by the first entity.

3 . The method of claim 1 , wherein the tree comprises decisions based on whether an improvement, occurring after choosing a pair based on the entities in the pair sharing a feature attribute, is above a threshold.

4 . A system for matching computer equipment or software entities using artificial intelligence, the system comprising:

a memory; and

a processor to:

determine, from a set of features describing entities, a subset of features which when input to a machine learning module allow the machine learning module to match a pair of entities based on the ability of the second entity in the pair to successfully take over responsibilities handled by the first entity in the pair, wherein the pair of entities have disjoint features, and wherein the determining the subset of features comprises using a decision tree model, the tree comprising: decisions based on whether an entity in a pair has a feature above a threshold;

periodically train a machine learning module using the determined features, wherein the machine learning module comprises a random forest model, and wherein training the machine learning module using the determined features comprises providing the machine learning module with pairs of entities and for each pair an amount of improvement occurring when the second entity in the pair took over the responsibilities of the first entity in the pair;

collect, using one or more microservices, information regarding a set of entities;

for a first entity described by the determined set of features, use the machine learning module to determine a list of entities matching the first entity, each entity in the list of entities associated with a matching score, wherein the determining of the list of entities matching the first entity comprises removing a feature from the input to the machine learning module and determining the effect on accuracy of the removing, and wherein the first entity described by the determined set of features and each entity in the list of entities each comprises computer equipment or software to be replaced; and

automatically activating an entity in the list of entities.

5 . The system of claim 4 , wherein the matching score for an entity reflects the ability of the entity to successfully take over responsibilities handled by the first entity.

6 . The system of claim 4 , wherein the tree comprises decisions based on whether an improvement, occurring after choosing a pair based on the entities in the pair sharing a feature attribute, is above a threshold.

7 . A method for determining the most successful pairing of computer equipment or software entities, the method comprising:

determining, from a set of features describing entities, features which when input to a machine learning module allow the machine learning module to pair entities based on the likelihood of the second entity in the pair to be successful when replacing the first entity in the pair, wherein the first entity in the pair and the second entity in the pair have disjoint features, and wherein the determining of the features comprises using a decision tree model, the tree comprising: decisions based on whether an entity in a pair has a feature above a threshold;

periodically training a machine learning module using the determined features, wherein the machine learning module comprises a random forest model, and wherein training the machine learning module using the determined features comprises providing the machine learning module with pairs of entities and for each pair an amount of improvement occurring when the second entity in the pair took over the responsibilities of the first entity in the pair;

collecting, using one or more microservices, information regarding a set of entities;

for a first entity described by the determined set of features, using the machine learning module to determine a list of entities to be paired with the first entity, each entity in the list of entities associated with a score, wherein the determining of the list of entities to be paired with the first entity comprises removing a feature from the input to the machine learning module and determining the effect on accuracy of the removing, and wherein the first entity described by the determined set of features and each entity in the list of entities each comprises computer equipment or software to be replaced; and

automatically activating an entity in the list of entities.

8 . The method of claim 7 , wherein the score for an entity reflects the ability of the entity to successfully take over responsibilities handled by the first entity.

9 . The method of claim 7 , wherein the tree comprises decisions based on whether an improvement, occurring after choosing a pair based on the entities in the pair sharing a feature attribute, is above a threshold.