Design a Product Recommendation Engine for a Telecom operator -- 3
Budget: $30 – $250 USD
Currently, marketing campaigns for Upsell/Cross-sell/Retention are based on simple business rules. The task is use of combination of business rules and machine learning to come up with an automated recommendation engine. The engine should operate at a segment of one (i.e. which means the campaign and product should be personalized for each customer based on their individual behavior and other patterns).
In order to build this tool, participants may use data inputs like voice and data usage, Age on Network (customer age with the operator), device/handset information, revenue data like monthly plan, payment/recharge information, product information like plan type, plan benefits, validity; call center and in store information related to complaints, queries and resolutions; customer demographics like age, income etc. any other variable/feature that they deem important for the engine.
Your Mission
In a team of 5-7 participants, design a AI-based Recommendation Engine to generate “next best action/product” recommendation for the Telcom customers. The Engine/tool will help the marketing team to decide a personalized consumer marketing strategy (Upsell/Cross-sell/Retain) and roll out offers in real time to maximize revenues.
Key Deliverables
Overall Approach and architecture of the Recommendation Engine
Explain and define the main components of the Engine? Will it take inputs from other Machine Learning Models?
Identification of appropriate strategy (Upsell/Cross-sell/Retention) for each customer using business rules + machine learning What are the typical key triggers that would be used for the same
How would the engine map the strategy with the product available in the product catalogue based on the customer behaviour?
How would you ensure Revenue cannibalization is minimized? How are the customers having high probability to churn treated in the model?
What additional data would have made the model better and how would that be integrated to the model?
Tools to be used for Designing the Recommendation Engine
Please use Python/R or any other programming language for coding. Dummy data to be used (can be taken from Kaggle or other data sources) and assumptions to be made.
submit it on 19/11 Morning 9:00 (UAE time)
In order to build this tool, participants may use data inputs like voice and data usage, Age on Network (customer age with the operator), device/handset information, revenue data like monthly plan, payment/recharge information, product information like plan type, plan benefits, validity; call center and in store information related to complaints, queries and resolutions; customer demographics like age, income etc. any other variable/feature that they deem important for the engine.
Your Mission
In a team of 5-7 participants, design a AI-based Recommendation Engine to generate “next best action/product” recommendation for the Telcom customers. The Engine/tool will help the marketing team to decide a personalized consumer marketing strategy (Upsell/Cross-sell/Retain) and roll out offers in real time to maximize revenues.
Key Deliverables
Overall Approach and architecture of the Recommendation Engine
Explain and define the main components of the Engine? Will it take inputs from other Machine Learning Models?
Identification of appropriate strategy (Upsell/Cross-sell/Retention) for each customer using business rules + machine learning What are the typical key triggers that would be used for the same
How would the engine map the strategy with the product available in the product catalogue based on the customer behaviour?
How would you ensure Revenue cannibalization is minimized? How are the customers having high probability to churn treated in the model?
What additional data would have made the model better and how would that be integrated to the model?
Tools to be used for Designing the Recommendation Engine
Please use Python/R or any other programming language for coding. Dummy data to be used (can be taken from Kaggle or other data sources) and assumptions to be made.
submit it on 19/11 Morning 9:00 (UAE time)