In the US, there are a lot of retail dealers that operate under various wireless carriers to drive Device & Service sales. These dealers operate independently but sell using tools offered by the carrier and some tools from 3rd parties. Sprint was one such carrier whose dealers faced a challenge – a wealth of underutilized data and recurring issues of chargebacks, lease frauds, and transactional discrepancies. We built a platform for a couple of such dealers that gave them actionable insights about frauds & discrepancies and a birds-eye view of their organizational sales.
Client Expectations
Improving the way transactions are processed both in terms of time and accuracy.
Flagging off transactions that can be potentially fraud or have discrepancies.
Making the analysis actionable, thus allowing users to make decisions and record those decisions in the system itself.
Giving everyone access to their transactions holistically.
Data Aggregation:
Fetch and consolidate data from multiple Sprint and Dealer-owned 3rd party portals.
Data Quality Enhancement:
Data may have certain missing elements that can set off the accuracy of our analysis. Develop rules and paradigms to address that.
Building a Robust Process:
Processes and data are usually volatile in this industry, so we had to ensure that our system was robust enough to handle changing requirements.
Accurate Commission Reconciliation:
Commission reconciliation is a challenging task for dealers and if we do it manually, it takes weeks to reconcile. Automation shouldn’t come at the cost of accuracy.
Building Actionable System:
Building a platform helping to generate actionable insights and track actions and remedies taken.
18 months of collaboration with the client
Product Development, Web Development, and Data Analytics & AI
Team Composition of 7 talented individuals who worked seamlessly together
Created a robust data engineering pipeline that included RPA and Python scripts to fetch data from the source and then ingest it into the data warehouse
Built data transformers that were needed to parse and process data in a better way to ensure proper conversion between changing formats
Created a rule-based AI (GOFAI) engine to flag discrepancies and frauds and assess refunds and chargebacks accordingly
Built a platform for auditors, management, and employees to track the transaction vetting process and to track actions taken on the insights generated by the GOFAI engine
Created a further process to verify fraudulent transactions by checking email validity by email bounce analysis and robocalling
Delivered valuable insights at all organizational levels
Achieved client cost savings exceeding one million dollars within 1.5 years
Enabled detection of both intentional and unintentional losses by employees
Our processes led to the detection of frauds & discrepancies in 4 days instead of an earlier 3-4 weeks
Want to bring clarity to your data governance, automated AI workflows or build smart systems to help with better decision-making?