CHALLENGES A leading tax authority in APAC was facing increased incidences of misreported income and VAT declarations from transactions conducted on e-commerce platforms, leading to underpaid taxes. It struggled to scale the processes of identifying audit candidates and conducting investigations, due to the massive volume and siloed nature of its data Multiple sources: Data from siloed platforms, tax submissions, and internal databases could not be fused to present all information pertaining to each individual or company as part of a holistic profile. This made calculations and tagging of audit candidates difficult and time consuming. Suspicious indicators: Without advanced analytical capabilities, the tax authority couldn’t effectively flag suspicious indicators and individuals and was thus unable to prioritize audits based on potential risk. Unstructured data: The growing trend of video streaming and photo sharing on online platforms for promotional and sales purposes creates difficulties in accurately identifying sales made across different online platforms. SOLUTION The tax authority turned to Cognyte and asked to use its decision intelligence platform, NEXYTE, with the goals of boosting tax collection related to online marketplace sales and improving the audit and investigation process. NEXYTE solution pillars: Data fusion: Using NEXYTE, the tax authority was able to fuse data from separate platforms, tax submissions, VAT records, and internal databases to form holistic profiles with all the information investigators needed for their work. This enabled the tax authority to uncover connections between suspicious profiles, detect fraudulent networks, and identify business relationships. Based on the fused information and analytical insights, the NEXYTE platform automatically calculated and tagged discrepancies in past and current tax documents and other sources. CHALLENGES • Prioritizing which individuals and businesses to audit • Discovering and investigating online sellers engaged in tax evasion • Correlating data between online marketplaces and tax declarations • Analyzing unstructured data, including video, to gauge sales SOLUTION • Fuse and compare marketplace seller data and tax forms • Leverage video analytics to estimate sales volume of goods and services offered on recorded or live video streams • Use ML-based risk scoring to effectively surface and prioritize audit candidates OUTCOME • Risk scoring accuracy of 85% allowed better prioritization of relevant audit candidates • Enhanced ability to detect fraudulent trends and correlate data to find serial tax evaders • Tax collection from e-commerce platforms more than doubled compared to the prior year Combatting tax evasion on online marketplaces with decision intelligence Tax authority case study