This isn't your fathers RA System

With apologies to GM (General Motors), based on our experience designing and delivering RA Systems to operators worldwide we are seeing how the RA landscape is changing. This evolution from the legacy to the emergent model is worth a once-over and that is the focus of this paper

When RA came into vogue in the early 2000s (late 90s ?) the charge was to analyze data from various elements in the operator’s revenue chain to identify leakage and then reduce or eliminate the problem. One of the key issues in RA was handling the voluminous data that operators generated - and this was before ‘Big Data’. So we had to manage big data before there was Big Data (Distributed Processing Technology). The technology, processes and features had to be fitted to handle this issue like sampling rather than full analysis and D-2 / D-3 (Processing day minus 2 or 3) instead of Real Time with the limited options we had. Because of the complexity of handling multiple data formats - ETL and data management were quite brittle. Developing and implementing RA systems were not a simple task and it was akin to an ERP implementation? - minus the notoriety

The organizational role was also murky at times - IT, but not fully IT, Finance, but not fully Finance. And also RA was not combined with FM (Fraud Management) as it would be later in its evolution. The focus was distorted as building the library of adapters and a framework for data acquisition typically for MSC, IN, HLR, Billing Systems, TAP, and CRM covering ASN.1, JSON, XML took priority over providing results to the business and finance team. And generally the posture was reactive. Because of the technological limitations the level of analysis that could be done was also limited and the ‘wins’ were not always clear. The mandate was to analyze the data, identify leakage, fix the leak, and save money. The big 3 4 were MSC v IN, Tap files, HLR analysis & sample rerating. The role was by comparison limited

The Wonder Years

RA systems started hitting its stride with the advent of Big Data - now the scope increased as operators, vendors and systems matured. RA departments became standard and even greenfield operations included RA. The types of analyses increased and RA reporting and KPI tracking became part of standard daily operations.?

Cloud computing that enabled Big Data was not fully leveraged by operators as pretty much all regulators across the board prohibited cloud installs for telcos citing security concerns. But the cloud native technologies were translated to on-prem installs also and with some limitations those advances were incorporated into RA architectures. Now terabytes of data could be processed daily, Hadoop allowed large scale batch processing, streaming data can be processed through advanced queuing and parallel dashboarding and visualization jumped ahead thanks to innovations in web technologies

Interestingly other verticals also started noticing this area of assurance , so now Finance, Insurance, Banking, E-Commerce were considering RA as another assurance operation that they should consider. While it will take another decade (2020s) before it will become actively explored in these domains

Back To the Future

The future is now - I think. RA and now RAFM has morphed into a Decision Support System or Data Driven Decision System. This is due to the breadth of reach of RA systems, the output from this department is now an important part of senior management decision making. And with great power comes great responsibility (had to say it). Now we see RAFM in better relief. The first significant change is the elevated mindshare from the operators. This in turn means that often RA systems by default become the source of truth. Secondly the scope of this assurance is now enhanced to include Margin analysis, Balance movement, IFRS and GAAP reporting. Additionally analytics scope has expanded Revenue Forecasting, Customer Churn Analysis, Customer Segmentation and Product Performance. RA can now provide Audit Ready reporting and perform audits as well. Margin analysis is a powerful tool that was not available to operators before allowing Marketing and Finance clearer visibility into financial performance

RA has emerged from the back office to the front office. Operators expanded RA use to include performance verification of MSCs, IN, Charging Systems, network elements etc.,

?This means RA systems themselves are packaged differently. Now the solution is integrated with data governance and data management features in the back end to ensure that the data quality is maintained as well as access to data sources is available. RA system can now maintain data catalog through data lake integration

The Good The Bad and The Ugly

Integrating data governance is a key differentiator. Cleaning up the data in the process - keeping the good, fixing the bad and discarding the ugly. Now we have the beginnings of a comprehensive decision support system. This marks a significant change in how RA systems work, how RA systems are viewed and how RA systems contribute to the organization. This is key because now RA is an enterprise effort across domains. Given this evolution, it’s time to think beyond Revenue Assurance (RA) and consider a broader concept—Data Integrity Assurance (DIA)

For instance, reconciliation between the MSC and IN ensures that the revenue forecast generated by IN reports is reliable. If we were to forecast revenue from IN without reconciling with MSC and instead relied solely on the data completeness from IN, the results might not be accurate. DIA reflects the expanded role these systems play, ensuring not just revenue but the integrity and reliability of all data, transactions, performance metrics, valuations, code, logic, and policies across the enterprise. Whether in telecom, banking, or fintech, DIA is becoming a critical component of the corporate application stack

The consensus is that there is ‘gold in them thar hills’ - well in them mountains of data anyway. This has moved RAFM to a new position in the corporate app stack to DIA. DIA is emerging as a? major focus of CTOs on its own -? as in a data driven world,? job number one is to make sure you have the right data and it is trustworthy

Integrating the data component is a natural progression of how assurance in general has morphed to what it is today - it is a data play. Because when you are in the business of assurance, you are by nature assuring everything you are touching. Be it data, transactions, performance, valuations, code, logic, policies etc. So a natural fit

A.I. Artificial Intelligence?

(A.I. Artificial Intelligence (2001) - IMDb). There is an inexorable pull into the next great thing. Everything will be A.I at some point. Resistance is futile. This is the end piece to the package. We have seen how RA has grown from a rudimentary analytical system? to a full spectrum data decision system with DIA. The next step is A.I. It is a natural fit and straight line progression. Today we can add a next level of review of the analyses by AI. The number of reports and KPIs that are generated by the RA system is reaching extra human levels - meaning it cannot be meaningfully managed by human analysts on a daily basis and needs a layer of automation to allow the humans to focus on the important issues that AI has identified

AI today can review KPIs and pick the anomalous ones, perform the first layer of verification and alert the human. This seems very logical and expected and eminently doable??

The Eternal Sunshine Of the Spotless Mind

We have come to the part where all loose ends are tied and good guys win. This has been a journey of success, the arc of the story has been one of good - bends towards triumph. Todays RA is truly a decision support system better than what it was envisioned. This is now a pan vertical solution, completely self-services supported by AI. Meeting the needs of the corporate assurance and teams that need data at large