SaaS Expense Matching via Machine Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SaaS management platforms face challenges in efficiently and accurately matching expense records to software purchases due to incomplete descriptions, similar software titles, and lack of access to disambiguating data sources, leading to low accuracy and high costs.
Innovation Solution
Implementing a SaaS management platform that uses machine learning models to analyze expense records, HR data, application engagement data, and other supplementary information to differentiate between software and non-software expenses, and identify specific software titles, leveraging APIs from various systems for data integration and processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional rule-based matching is used to match expense records to software, then the system is simple to implement, but the matching accuracy is low due to incomplete descriptions and similar software titles
Solution Approach 1:
The patent replaces traditional rule-based mechanical matching systems with machine learning models that automatically learn patterns from data. The ML models analyze expense records, software titles, and additional data sources to identify matches without relying on pre-defined rules, thereby improving accuracy while managing complexity through automated learning.
Solution Approach 2:
The patent introduces additional data sources (HR data, application engagement data, SSO data) as intermediaries to bridge the gap between expense records and software titles. These intermediate data sources provide contextual information that helps disambiguate similar software titles and improve matching accuracy.
2Measurement precision
If multiple data sources are integrated to improve matching accuracy, then the matching precision improves, but the resource intensity and processing time increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing and indexing data from multiple sources before matching occurs. HR data, application engagement data, and SSO data are collected and structured in advance, allowing the ML models to efficiently query and compare information during the matching process without excessive computational overhead.
Solution Approach 2:
The patent segments the matching process into distinct stages: collecting expense record data, collecting additional contextual data from multiple sources, processing and analyzing the combined data using ML models, and generating matches. This segmentation allows each stage to be optimized independently, managing resource intensity effectively.
3Reliability
If manual review processes are used to verify expense matches, then the accuracy of software identification improves, but the processing time and operational costs increase
Solution Approach 1:
The patent implements a self-service system where the ML models automatically perform the verification and matching functions that would traditionally require manual review. The system autonomously analyzes expense records against multiple data sources and identifies software purchases without human intervention, maintaining high accuracy while dramatically improving processing throughput.
Solution Approach 2:
The patent incorporates feedback mechanisms where the ML models continuously learn from match outcomes and can be refined based on verification results. This feedback loop allows the system to improve its accuracy over time while maintaining automated high-speed processing, eliminating the need for continuous manual review.
Data Source
AI summary
A Software as a Service (SaaS) management platform (SMP) is provided, the SMP configured to perform a method including: receiving a plurality of expense records for a customer organization, wherein each one of the plurality of expense records identifies a user, a vendor, and an amount spent; using a first machine learning model to analyze the plurality of expense records and determine which ones of the expense records represent software purchases; using a second machine learning model to analyze the expense records that are determined to represent software purchases, and identify software titles that the software purchases are for; surfacing the expense records that represent software purchases, in association with their respective identified software titles, through a user interface of the SMP for the customer organization.