Project Detail
Contact Optimization for Improved Customer Interactions
Built predictive models and recommendation logic to improve customer interaction strategies while refactoring an existing application for better maintainability and performance.
- Role
- Data Scientist
- Context
- Telecommunications
- Domain
- AI Applications
- Duration
- 15 months
- Status
- Finished
What changed
Developed machine learning models to predict hotline demand and support more proactive customer engagement.
Designed recommendation logic that helped call center agents identify more suitable customer tariffs, improving call efficiency and customer experience.
Refactored the application codebase to improve maintainability, readability, and performance.
Integrated model outputs into customer service workflows, enabling day-to-day operational use of predictions.
Technology
Python
6 technologies
pandas
seaborn
PySpark
pytest
Black
Ruff
Machine Learning
2 technologies
Random Forest
XGBoost
Databricks
5 technologies
Unity Catalog
Databricks Clusters
SQL Warehouses
Databricks Asset Bundles
MLflow
SQL
2 technologies
PL/SQL
Spark SQL
Signals
machine-learning
data-science
customer-analytics
application-development
mlops