Customer 360 Analytics
- The client wanted us to help them with an efficient mechanism to get the complete customer view in a mall.
- The system must be capable to predict the user movement and shopping pattern of each customer who would swing by the mall.
- To map the path taken by a user without using any beacons.
- Predict next store to visit using Deep learning (TensorFlow).
- The malls can get a track on the users path per day and their activity using their device and MAC ID.
- Recommend users to stores and stores to users based on collaborative filtering.
- To exactly predict the most possible path taken by the visitor between two different locations.
- Custom algorithm which uses nearest neighbour logic to identify the most possible path taken by two different locations was implemented for better accuracy.
- A custom web dashboard that shows the density plot of visitors in the mall. The malls can get a track on the users path per day and their activity using their device and MAC ID.
- The malls can decide on the shop rents based on the traffic to specific areas.
- Apache Spark
- Apache Hive
- AWS S3