Vehicle Deal Microservices Parallelization for Low-Latency Pricing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern computing environments struggle to compute multidimensional vehicle deals in a timely manner due to latency issues in distributed microservices architectures, failing to meet the one-second turnaround time expected by prospective vehicle purchasers.
Innovation Solution
A double parallelization approach combined with a succinct, non-duplicative data structure is employed to optimize data transfer and computation across microservices, ensuring parallel execution and re-aggregation of results in the correct order.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If distributed microservices architecture is used, then system scalability and modularity are improved, but computation latency increases and fails to meet one-second turnaround time requirement
Solution Approach 1:
The patent segments the computation task into two independent parallel streams: risk score computation and pricing computation. Each stream processes different aspects of the vehicle deal evaluation independently, allowing simultaneous execution without sequential dependencies. This segmentation eliminates the time loss inherent in sequential microservice calls while maintaining the modular architecture benefits.
Solution Approach 2:
The patent introduces a new dimensional approach by implementing double parallelization - parallelizing both the computation streams (risk vs. pricing) and the data transfer operations (pushing data proactively before computation completes). This multi-dimensional parallelism transforms the single-threaded sequential processing into a multi-threaded concurrent system, achieving sub-second response times while preserving microservices scalability.
2Productivity
If data is transferred between microservices, then computation can be distributed, but data duplication and transfer overhead increase
Solution Approach 1:
The patent applies preliminary action by proactively pushing computation data from the data layer to the computation services before the actual computation is needed. The system pre-transfers the vehicle data, user data, and deal data to both the risk score service and pricing service in advance, so that when computation is triggered, the data is already available locally, eliminating redundant transfers and reducing data duplication overhead.
Solution Approach 2:
The patent creates a universal data structure that serves multiple functions: it acts as both the input data for risk score computation and the input data for pricing computation. By designing a multi-functional data transfer format that contains all necessary fields for both computation types, the system eliminates the need for separate data transfers and duplications, achieving efficient distributed computation with minimal data redundancy.
3Reliability
If sequential processing is used, then data integrity and correct ordering are maintained, but computation speed is too slow to meet user expectations
Solution Approach 1:
The patent segments the sequential processing into parallel computation streams while maintaining data integrity through separate processing paths. The risk score computation and pricing computation each receive identical input data but process it independently through their own logic, ensuring that parallelization does not compromise data integrity. The segmentation allows speed improvement while preserving reliability.
Solution Approach 2:
The patent implements feedback mechanisms to ensure correct ordering and data integrity in the parallel processing system. The system uses correlation IDs and result aggregation logic that feedback on the completion status of each parallel stream, ensuring results are combined in the correct order and that data integrity is maintained throughout the parallel computation process.
Data Source
AI summary
A computer-implemented approach is described that includes receiving a multidimensional vehicle-deal request for computation of a plurality of vehicle-deals. To overcome latency and scalability issues, the approach computes the results for the multidimensional vehicle-deal request using double parallelization. Such double parallelization involves computing pre-price and post-price calculations using two asynchronous parallel threads using microservices servers. The results are output on the device of the requester, optionally in an order sorted based on most attractive results to the user.