Data Analyst Portfolio: SQL Analysis, Tableau Visualization, and Machine Learning Project Practice

Siti Suharyanti's GitHub portfolio covers three projects: SQL analysis of the Chinook music store, Tableau revenue visualization dashboard, and tweet sentiment analysis for PT Esteh Indonesia Makmur. It demonstrates the complete data science process from data extraction to visualization, making it an excellent reference case for junior data analysts seeking jobs and learning.

• Original author/maintainer: Siti Suharyanti (aspiring junior data analyst)
• Source platform: GitHub, project title Data-Analyst-Portfolio, link https://github.com/SitiSuharyanti/Data-Analyst-Portfolio, published on June 4, 2026
• Portfolio positioning: Demonstrates capabilities in SQL querying, data visualization, machine learning, etc., covering the complete process from data extraction, cleaning, analysis to visualization.

Project Methods

1. SQL Analysis: Uses PostgreSQL to process the Chinook database, covering skills like aggregate functions, table joins, CTEs, window functions, etc.

2. Tableau Visualization: Presents revenue trends, customer segmentation, etc., via interactive dashboards, converting SQL results into intuitive visualizations.

3. Machine Learning: Uses Naive Bayes and SVM to classify Twitter sentiment, including data scraping, preprocessing, model building, and evaluation.

Tech Stack

• Database: PostgreSQL
• Language: Python
• Visualization: Tableau, matplotlib, seaborn
• NLP: Sastrawi, demoji
• Data Processing: pandas, numpy
• ML: scikit-learn (Naive Bayes, SVM), SMOTE, TF-IDF
• Version Control: Git/GitHub

SQL Project

• Dataset: 59 customers, 412 invoices, 3503 tracks, covering 24 countries, from 2021 to 2025
• Tools: PostgreSQL, pgAdmin

Tableau Project

• Outcomes: 2 interactive dashboards, meeting the needs of executives (revenue overview) and operations (track performance)

Machine Learning Project

• Dataset: Tweets containing "es teh indonesia" and "somasi" from September 24-30, 2022, scraped via snscrape
• Tools: Python, Google Colab, scikit-learn
• Skills: NLP preprocessing, TF-IDF, SMOTE, model evaluation (accuracy, F1 score, etc.)

Highlights

• Complete technical loop: SQL extraction → Python processing → Visualization presentation
• Domain-specific: Indonesian NLP processing, addressing non-English text issues
• Project strategy: Choosing commonly used datasets in education (Chinook) and hot events (Twitter sentiment)

Learning Value

• Provides a clear learning path: SQL basics → Python data processing → ML and visualization
• Project-driven learning: Each project has clear business scenarios and technical goals, which is more effective than grammar learning alone.

1. Projects cover core links: Data extraction (SQL), analysis and modeling (Python/ML), result presentation (visualization)
2. Clear project descriptions: Each project should specify business background and technical details to reflect value
3. Intuitive skill display: List tools and technologies in categorized tables to facilitate recruiters' demand matching
4. Provide accessible links: Point to code repositories or online demos for in-depth understanding