How can Big Data help in the above transformation? By definition, Big Data is a collection of data that is huge in volume, yet growing exponentially with time. It is data with intricate complexity that traditional data management tools cannot store or process efficiently. Big Data processing therefore requires advanced analytics, driven by machine learning, to churn and yield effective output for efficient wrangling by its custodians.

The GSTN is a perfect example of how Big Data is stored and processed/ analyzed in data lakes to give insights into an ecosystem. The GSTN environment grows daily with almost 1.03 crore taxpayers and their data presently integrated in the system. The revenue authorities in India have been taken the initiative to wrangle this data on two fronts. First, being the effective use of data to bring about ease of doing business; while the second, more complex part, being intelligent data analysis for risk management and protection of the ecosystem. In the GSTN, the authorities, to manage the perspective of risk management, are now using advanced analytics and data mining tools to conduct supply chain analysis of data garnered from registered taxpayers. This is helping the authorities to identify outliers, like risky major suppliers, whose filings show deviations from established patterns. Once a pattern emerges after the analysis of a target supplier, the data is shared with jurisdictional field officers or revenue intelligence authorities to conduct further investigations and identify the glitch in the business.

The analysis of Big Data in GSTN started from early 2019. However, this exercise already had a growing predecessor in the Business Intelligence Analytics Module- ‘ÁDVAIT’, which was put in place by the Indian Customs authorities in early 2018. The ADVAIT system is a complex network of machine learning which integrates data from transactions done in Customs ports, during the course of Exim trade, with data from other intelligence systems (e.g. Income tax filings, the Reserve Bank PFMS system, data from Directorate General of Foreign Trade etc.) and builds a profile of a business engaged in Exim trade. Once such profile is built, the same is developed and nurtured through complex algorithms, built on the basis of established law and business dynamics, to develop an evolving profile, which can be accessed by revenue intelligence at any time if an anomaly in trading pattern is suspected. Based on the business intelligence generated by ADVAIT, Exim trade is also extended facilitated trade clearances at Customs ports (example: the Authorized Economic Operator Program or the Direct Port Clearance Program). This also helps in bringing down dwell time in the clearance of Customs goods bringing relief to the economics of the supply chain of the stakeholder.

The ADVAIT modules are also supplemented to a large extent by the intelligent initiatives introduced by the Indian Customs to bring about Ease of Doing Business including but not limited to the introduction of system driven ‘faceless customs clearances’ at Customs ports, the improved Risk Management System (RMS 2.0) which facilitates expedited clearance of Exim goods based on established databases and clearance patterns, and other intelligent modules such as the Indian Customs EDI system (ICES 5.1), which compiles databases from all filings made by an Exim business at Customs ports. While, the overall dynamics of the machine are not free from teething glitches, the experience of the Exim community has largely been satisfactory after the introduction of these new facilitation measure based on business intelligence mining by the Indian Customs.

Coming back to GSTN, how is the learning generated by business intelligence techniques of Customs being effective used by the revenue authorities? On the administration part, new electronic frontiers are being introduced in the system which serve the dual purpose of facilitating trade and the electronic sweep of information/ data for the algorithms to learn and establish patterns. This include the introduction of auto-generated input credit statements (purportedly to help taxpayers efficiently pay taxes, match and file returns); electronic filing of invoices with QR codes; Aadhar based authentication for easier registration etc. On the enforcement front, data analytics is helping identify bogus or fraudulent elements in the system (like credit dodgers and spurious refund claims), fake registrations and other such unscrupulous operators, who would have otherwise be lost in the massive swirl of data generated in the system. To wit, revenue authorities purportedly detected INR 37,946 crore worth of tax fraud in FY 2018-19 and INR 6,520 crore within the data filed in the April-June period of 2019-20. Elements of fake involving INR 11,251 crore, were also detected, leading to the arrest of 150 persons. The above enforcement was largely credit to advanced analytics introduced by the Government.

In a nutshell, machine learning and Big Data analytics is the ecosystem’s ability to see hidden or emerging patterns in existing data. These patterns help improve performance and increase the efficiency of the electronic ecosystem. Analytical tools and predictive modelling now enable revenue authorities to study historical data and compare the same with current business practices and the use of predictive modelling is akin to identifying ‘tells’ in a card game, which may be invisible to the naked eye but be snared from a machine learning perspective. While, the use of business intelligence has its share of benefits, care should also be taken not to be reliant only on the system, as sometimes false positives can also be generated. The idea of the day should be the effective use of hands-on administration coupled with the electronic data analysis in order to bring about a sustainable tax environment.

Read Less-