Practical SQL, Excel, Python, and Power BI analytics for questions that need a clear answer, not just a chart. Based in Lagos, working with data wherever it lives.
I own analytics projects end to end — from a messy raw dataset through to a dashboard a non-technical stakeholder can open and actually understand.
Joins, subqueries, and window functions against PostgreSQL and SQLite, structured into clean star schemas built for reporting.
Interactive Power BI dashboards with DAX and proper data modeling, Tableau, Excel, and Streamlit. Built for executives and non-technical audiences, not just other analysts.
Time-series decomposition, statistical anomaly detection, and forecast models validated against real holdout data before anyone trusts the number.
Each project below went from raw data through SQL and Python to a dashboard or app a stakeholder actually opens — built at Infotact Solutions & Co.
Replaced last-click attribution bias by merging ad spend, web analytics, and CRM conversion data with SQL window functions, sequencing each customer journey, then building a Power BI dashboard that lets managers toggle between First-Touch, Last-Touch, and Linear models.
View on GitHub →Grouped users by acquisition month in Python and PostgreSQL to identify exactly when churn happens, built a retention heatmap, and calculated 12-month Customer Lifetime Value by segment — delivered as a four-page Power BI dashboard and a published LinkedIn article.
View on GitHub →Decomposed daily sales into trend, seasonality, and residuals, benchmarked Z-score, IQR, and Isolation Forest anomaly detection against decomposition residuals to cut false positives, then validated ARIMA and Prophet forecasts with chronological train/test splits — shipped as an interactive Streamlit app.
View on GitHub →Every project I ship follows the same process, whether it's a two-day dashboard fix or a four-week forecasting pipeline.
Before writing a single query, I get clear on what decision the analysis actually needs to support — and who's making it.
Missing values, inconsistent timestamps, duplicate records — resolved and documented, with assumptions stated up front, not buried.
Trend, seasonality, and residuals get looked at before I trust any pattern — a spike is often seasonality, not a signal.
Multiple methods benchmarked against each other, validated on real holdout data — the model that wins on paper has to prove it first.
Findings become a Power BI dashboard or an interactive app a non-technical stakeholder can open and use without me in the room.
Assumptions, limitations, and reasoning get written down — version-controlled on GitHub, not left in my head.
A data analyst who owns projects end to end — from writing SQL against relational databases to building Power BI dashboards non-technical stakeholders actually use. Strong foundation in financial modeling and KPI reporting, with a track record of communicating findings to executive audiences.
The best dashboard is the one nobody has to ask you to explain.
I'm a Data Analytics Expert at Infotact Solutions & Co., where I've owned three end-to-end analytics projects — marketing attribution, customer retention, and supply chain forecasting — from raw data through stakeholder-ready dashboards. Before that, I built financial and operational dashboards as a Data Analyst Intern at Bliss Analytics.
I hold a B.Eng. in Mechanical Engineering from Obafemi Awolowo University, and I build AI tools into my daily workflow to move faster on SQL, Python, and documentation — while always validating and correcting the output myself before it reaches final analysis.
Open to Data Analyst and Power BI Developer roles. If you've got a messy dataset and a real question behind it, I'd like to hear about it.