👋 Hi, I’m Frank

Data Analyst · Data Scientist · Data Engineer
Power BI | SQL | Python | Databases | ETL | Forecasting

I’m a passionate data enthusiast who bridges analytics, data science, and data engineering.
I turn raw data into business intelligence, predictive insights, and automated data pipelines, furthermore, constantly exploring how analytics, machine learning, and data infrastructure can work together to create real impact.

---

🔧 Core Skills

• Analytics & BI: Power BI (DAX, Power Query), KPI dashboards, executive reporting
• Data Science: Python (pandas, NumPy, scikit-learn, statsmodels, Prophet), EDA, modeling, forecasting
• SQL & Databases: T-SQL, query optimization, CTEs/views, normalization, data warehouse modeling
• Data Engineering (in progress): ETL/ELT design, schema optimization, orchestration, data testing

---

🧰 Tech Stack

Languages: Python, SQL, DAX
Python Libraries: pandas, numpy, scikit-learn, statsmodels, prophet, matplotlib
BI & Visualization: Power BI (DAX, Power Query)
Databases: SQL Server, PostgreSQL, CSV/Parquet
Tools & Workflow: Jupyter, VS Code, Git, Power BI Service

---

📌 Highlight Projects

1️⃣ Liquor Business Analytics (Power BI + SQL)

Comprehensive multi-store liquor analytics for wholesale and retail.

• Designed normalized SQL schema and reusable analytical views
• Created DAX measures for cost variance, growth, and scan rate tracking
• Delivered live Power BI dashboards for store and category performance
  Tech: SQL Server, Power BI, DAX, Power Query, Python (data prep)

---

2️⃣ Demand Forecasting for Fast-Moving Products

Forecasted weekly sales for pallet purchasing and inventory planning.

• Applied Holt-Winters, SARIMA, and Prophet time-series models
• Detected seasonality and filtered refund-related outliers
  Tech: Python (pandas, statsmodels, prophet), Power BI

---

3️⃣ Deep Learning for Sentiment Image Classification

Built a CNN pipeline to classify image sentiment from visual cues.

• Modularized data loading, augmentation, and model evaluation
• Logged performance metrics and visualization
  Tech: Python, TensorFlow/PyTorch

Please be advised that the results are approximate and depend on runtime environments.

---

4️⃣ Banking & Pizza BI Dashboards

End-to-end Power BI dashboards analyzing customer behavior and performance.

• Data transformation with Power Query and semantic modeling
• DAX-based KPIs and trend visuals for management reporting
  Tech: Power BI, SQL, DAX

---

📈 Selected Achievements

• Reduced KPI reporting time from hours to minutes with reusable Power BI models
• Increased seasonal demand forecast accuracy through model tuning and validation
• Standardized data cleaning and timestamp parsing across multiple store datasets