Dynamic traffic pattern analysis and rate-limit adjustment

A method for application programming interface monitoring, the method comprising:

• receiving, for a plurality of application programming interfaces, usage monitoring data, wherein each application programming interface of the plurality of application programming interfaces is associated with a corresponding rate limit;
• inputting the usage monitoring data into a machine learning model to obtain a time period when a corresponding usage rate for one or more application programming interfaces is predicted to exceed the corresponding rate limit, wherein the machine learning model has been trained, using training data, to predict time periods when usage rates exceed rate limits;
• determining that the time period has been output by the machine learning model for an oversubscribed application programming interface, wherein the time period indicates that the oversubscribed application programming interface is predicted to exceed the corresponding rate limit;
• based on determining that the time period has been output by the machine learning model for the oversubscribed application programming interface, retrieving resource allocation data for the plurality of application programming interfaces;
• determining, based on the resource allocation data, one or more undersubscribed application programming interfaces from which one or more resources may be reallocated for the oversubscribed application programming interface;
• reallocating the one or more resources to the oversubscribed application programming interface by decreasing the rate limits for the one or more undersubscribed application programming interfaces for the time period;
• and
• increasing a rate limit for the oversubscribed application programming interface.

One or more non-transitory computer-readable media comprising instructions recorded thereon that, when executed by one or more processors, cause operations for monitoring application programming interfaces at a network system, comprising:

• receiving, for a plurality of application programming interfaces, usage monitoring data, wherein each application programming interface of the plurality of application programming interfaces is associated with a corresponding rate limit;
• inputting the usage monitoring data into a machine learning model to obtain a time period when a corresponding usage rate for one or more application programming interfaces is predicted to exceed the corresponding rate limit, wherein the machine learning model has been trained, using training data, to predict time periods when usage rates exceed rate limits;
• determining that the time period has been output by the machine learning model for an oversubscribed application programming interface, wherein the time period indicates that the oversubscribed application programming interface is predicted to exceed the corresponding rate limit;
• based on determining that the time period has been output by the machine learning model for the oversubscribed application programming interface, retrieving resource allocation data for the plurality of application programming interfaces;
• determining, based on the resource allocation data, one or more undersubscribed application programming interfaces from which one or more resources may be reallocated for the oversubscribed application programming interface;
• reallocating the one or more resources to the oversubscribed application programming interface by decreasing the rate limits for the one or more undersubscribed application programming interfaces for the time period;
• and
• increasing a rate limit for the oversubscribed application programming interface.