Access Bank
E- Channel Analytics
Shows the e-business team i sights into customer activations, transactions, churn, transactions patterns such as frequency or declining transactions across the bank electronic banking channels,
Designed the data mart for the analysis because the source data was coming from several databases (MS SQL, Oracle, MySQL) and also created the business requirement document used while reviewing dashboards designed by the data analyst.
TEAM LEAD
2018-2019
Customer Data Quality Management
This project is to help the bank access the quality of their customer data to help with marketing and communication needs and also to help to ensure compliance with regulations. The project aim to help clean customer input error, use business rules and external data to enrich customer data. They also want the system to flag exemption of customer who information do not meet the requirement of their account class as required by the CBN
Engaged business stakeholders to identify data domains, data quality business rule, data enrichment needs as well as data cleansing requirement and a formula for computing the bank data quality index. Design the Data Quality Mart for storing both valid and invalid customer records based on agreed rules and also help the analyst team to create data quality report and a live dashboard to see real time improvement and alert for data stewards
TEAM LEAD
2018-2019
Eco Bank
Advanced Analytics on Azure (Customer Segmentation and Customer Churn)
This project aim at using machine learning to help classify the bank customers into different segment to help improve profitability as well as identify customer at risk of churn so that account managers can reach out to them and prevent eventual churn
Worked on the data engineering side of the product by extracting data from the bank onprem source to azure database and performing necessary data cleansing and data aggregation on the data. Also work on the presentation module by designing dashboard that shows user the business users information on customers at churn risk also relevant metrics and recommendation dashboard on the segment distribution of their users
TEAM MEMBER
2019-2020
AFRICAN PRUDENTIAL
Cloud Based Enterprise Data Warehouse with Reporting
The business users want to be able to see reports on key business metrics across their different business units, they also want to be able to target certain users for marketing based on some filters.
The data warehouse project fulfilled this by ensuring that they have a single source of trust for enterprise data and the report created in power bi was embedded within the business website for some use cases and other connect to power bi for their report consumption
- Led the team to conduct requirement gathering with stakeholders(business and technical)
- Design the dimensional model
- Design the visualization template for reporting
- Design the Data Pipelines for Data Extraction, Transformation and Loading into the EDW while ensuring change data capture best practices are implemented
TEAM LEAD
2020
Tolu Elumelu Foundation
Data Quality Management
The Patners Database has several data quality issues due to data entry errors, the goal is to cleanse the data based on agreed rules and flag records that do not meet the data quality criteria so data stewards can reach out to the partners to fil the gap
TEAM LEAD
2020
United Bank for Africa
Azure-Based Advance Analytics covering customer churn, customer segmentation, product recommendation system, loan probability of default
The project also saved the bank about 400bn in deposit due to the accuracy in tracking churn and also help the risk in prioritizing loan portfolio and managing exposure to default
- Design the feature hub that extract data relative for analytics from the bank on premise data warehouse to azure data lake
- Wrote spark jobs that transform data from in the data lake into features for ML
- Perform Exploratory Data Analysis on the Data
- Improve feature by engineering new features based on data attributes and patterns
- Build and Test ML Models
- Deploy Models to Production
- Design ML Ops to ensure continuous model training and ensure that the best model are used for predictions
- Design Visualizations and report to show model metrics to end users ( the bank technical team) and business metric to business users
TEAM LEAD
2021-2022