Projects
This project has been built to extract customer reviews based on the trending campaigns in the market with respect to each car model & create an Online Brand Score with a deep-dive into its liked & disliked features for each model as well as provide the end user with a Competitor Analysis.
This model helps in detecting >90%+ promo code abusers at both registration and transaction level which replaced the legacy manual method used by my client thus reducing 95% of the time in identifying promo abusers.
The end results on my client database : for every 100 records, I achieved 94.55% Accuracy & 0.01% False Positives.
This project summarizes the sales performance of a Supermarket using Microsoft Power BI with a deep dive on various segments such as mode of payment, product-wise sales, type of sales, etc.
Anomaly Detection Framework
This enabled all the scorecard report owners in my team to use this framework on the datasets which are being fed in to their reports/dashboards to check for any bad data, anomalous behaviour, interesting insights on the KPIs in their data & provide the end user with automated mail alerts for the same.
