High-throughput, low-latency transaction processing
A method for low latency, high volume transaction processing. A set of transaction requests are received. In response to receipt of a transaction request, and for each transaction request, a set of tasks are orchestrated across a set of microservices, each of the microservices having a particular function uniquely associated with a transaction process, and wherein at least one microservice has an associated rules database. The database supports computation rules that are generated by applying natural language processing (NLP) to a rule set. Using the microservices, the set of transaction requests are concurrently processed, substantially in real-time, and by applying the set of computation rules. The results generated from concurrently processing the set of transaction requests are then returned. In one embodiment, each of the transaction requests are requests associated with an airline ticket and, in particular, a request to refund that ticket and any associated taxes and fees.
1. A method, comprising:
executing computing-based operations in a computing infrastructure comprising computer hardware and computer software, the computer software comprising a set of microservices having an associated rules database, each of the microservices in the set of microservices having a particular function uniquely associated with a transaction process, wherein at least one particular microservice of the set of microservices is a computer network-accessible software unit deployed in a dedicated process in the computing infrastructure and that comprises a search function and a discrete-time system, wherein executing the computing-based operations comprises:
generating a set of computation rules by applying machine-based Natural Language Processing (NLP) to a rule set;
hosting an instance of the set of computation rules in computer memory in the computer hardware;
receiving a set of transaction requests;
responsive to receipt of a transaction request, and for each transaction request, orchestrating a set of tasks across the set of microservices;
using the set of tasks orchestrated for each transaction request, concurrently processing the set of transaction requests in the set of microservices, in real-time, wherein one or more of the transaction requests are processed at least in part by the particular microservice (i) applying the search function against the set of computation rules hosted in the computer memory to generate a data set, and (ii) thereafter applying the discrete-time system against the data set; and
returning results generated from concurrently processing the set of transaction requests.
2. The method as described in claim 1 , wherein each of the transaction requests are requests associated with a ticket.
3. The method as described in claim 2 , and for a particular ticket, determining whether to initiate a refund associated with the ticket, wherein the refund includes one or more values generated by using the computation rules.
4. The method as described in claim 3 , further including providing control signaling to one or more third party systems to initiate settlement of the refund.
5. The method as described in claim 1 , wherein the set of tasks are orchestrated using a central controller.
6. The method as described in claim 1 wherein communications to and from the central controller are carried out via a publish and subscribe mechanism.
7. The method as described in claim 1 , further including authenticating and authorizing the request prior to orchestrating the set of tasks.
8. The method as described in claim 1 , wherein applying the search function against the set of computation rules includes identifying one or more segments or other identifiers associated with the transaction request, and using the set of computation rules to determine one or more values of the one or more segments or other identifiers.
9. The method as described in claim 8 , wherein the transaction request is a request for a refund of an airline ticket, the one or more values are taxes associated with a travel segment or coupon, and wherein the particular microservice is a tax calculation microservice that executes the search function and a non-linear polynomial discrete-time system to identify a tax associated with each travel segment or coupon.
10. The method as described in claim 1 , wherein the set of transaction requests are received programmatically from one or more third party computing systems.
11. The method as described in claim 1 , wherein the method provides a network-accessible service for airline ticket refunds.
12. The method as described in claim 1 , wherein a microservice is executed in a container.
13. The method as described in claim 1 , further including receiving from one or more third party computing systems given data that is at least one of: pricing structures, policy exceptions and other client-specific parameters, and using the given data to generate the result for a given transaction request.
14. A computing platform, comprising:
computing hardware;
computer software executing on the computer hardware, the computer software comprising a set of microservices having an associated rules database, each of the microservices in the set of microservices having a particular function uniquely associated with a transaction process, wherein at least one particular microservice of the set of microservices is a computer network-accessible software unit deployed in a dedicated process in the computing platform and that comprises a search function and a discrete-time system, the computer software comprising computer program instructions executed on the computing hardware to perform computing-based operations comprising:
generating a set of computation rules by applying machine-based Natural Language Processing (NLP) to a rule set;
hosting the set of computation rules in computer memory in the computer hardware;
receiving a set of transaction requests;
responsive to receipt of a transaction request, and for each transaction request, orchestrating a set of tasks across the set of microservices;
using the set of tasks orchestrated for each transaction request, concurrently processing the set of transaction requests in the set of microservices, in real-time, wherein one or more of the transaction requests are processed at least in part by the particular microservice (i) applying the search function against the set of computation rules hosted in the computer memory to generate a data set, and (ii) thereafter applying the discrete-time system against the data set; and
return results generated from concurrently processing the set of transaction requests.
15. The computing platform as described in claim 14 , wherein each of the transaction requests are requests associated with an airline ticket.
16. The computing platform as described in claim 15 , wherein the computer program instructions are further executed on the computer hardware to determine whether to initiate a refund associated with the airline ticket, wherein the refund includes one or more values generated by using the computation rules.
17. The computing platform as described in claim 16 , wherein the computer program instructions are further executed on the computer hardware to provide control signaling to one or more third party systems to initiate settlement of the refund.
18. The computing platform as described in claim 14 , operated as a network-accessible service.
19. The computing platform as described in claim 14 , wherein the computer program instructions executed on the computer hardware to apply the search function against the set of computation rules includes computer program instructions that identify one or more segments or other identifiers associated with the transaction request, and use the set of computation rules to determine one or more values of the one or more segments or other identifiers.
20. The computing platform as described in claim 19 , wherein the transaction request is a request for a refund of an airline ticket, the one or more values are taxes associated with a travel segment or coupon, and wherein the particular microservice is a tax calculation microservice that executes the search function and a non-linear polynomial discrete-time system to identify a tax associated with each travel segment or coupon.