Building Data-Driven Client Relationship Management in Banking with Python
summary
A bank replaces Excel sheets with a production Python stack that extracts features via SQLAlchemy-powered feature extraction, processes up to a million transactions per day with Luigi for million/day ETL, trains imbalanced classifiers with scikit-learn and imbalance-learn, and uses Random Forests plus treeinterpreter for per-lead rationales to produce ranked leads. A Django metadata browser, spaCy text topics, daily pipelines and a front-end workbench close the loop with monitored wins/losses and control-group evaluation.
AI-written summary · about
chapters (18)
- 0:00Intro
- 0:31Advanced Analytics in Banking
- 1:25Relationship Managers
- 3:30Excel Spreadsheets
- 4:10Client Planning
- 6:13Django
- 8:05Feature Extraction
- 8:22Sequel Alchemy
- 10:40Spending Income Profiles
- 11:43Text Data
- 14:29Dynamic vs Static
- 15:36Python Libraries
- 16:42Pipelines
- 18:53Machine Learning
- 19:45Making Recommendations
- 24:26Tracking Leads
- 25:02Summary
- 26:05Questions
related talks
35mBank evolution to data science-driven business model
40mCreating a Contemporary Lending Risk Management System Using Python
29mAI-Driven Business Success: Measurable Results and Revolutionized Sales
49mBuilding a Data-Driven Bank
02h 00mData Science Driven Business Insights with Python and PySp
02h 03mFull stack AI design and development
36mDigital, data-driven financial advisory
33mSmart Banking - Real Time Driven