Portfolio

Alyani Septalia

I am currently working on several self-directed projects in pursuit of formal work as a Data Analyst.

Education
UNMUL
Bachelor of Computer Science Universitas Mulawarman 2020–2026
Rakamin
Data Science Bootcamp Rakamin Academy 2023
2+ YoE in Data
Connect
Certifications
  • Microsoft
    Power BI Data Analyst Associate Microsoft · 2024
Skills

The Stack

Python Python
SQL SQL
Power BI Power BI
Tableau Tableau
Looker Looker
Jupyter Notebook Jupyter Notebook
Excel Excel
Git Git Notion Notion Figma Figma

Recent Projects

Global Plastic Waste Analytics
Python Python Jupyter Notebook Jupyter Notebook SQL SQL Power BI Power BI
2026

Global Plastic Waste Analytics

Public plastic waste data is fragmented with 7+ sources with inconsistent country naming, mixed granularity, and misaligned time coverage. I built an end-to-end pipeline to clean, harmonize, and merge them into a single analytics-ready master dataset, then published an interactive Power BI dashboard to answer questions like: which countries mismanage the most plastic per capita, how global production has trended over decades, and how waste fate differs across regions. Insights relevant for environmental policy and NGO reporting.

View Repository ↗ View Dashboard ↗
British Airways Lounge Eligibility & Booking Prediction
Python Python Jupyter Notebook Jupyter Notebook SQL SQL
2026

British Airways Lounge Eligibility & Booking Prediction

Airlines need to plan lounge capacity and improve booking conversion, but both require data they don't always model explicitly. Using a British Airways simulation dataset of 50,000 bookings, I built two solutions: a lounge eligibility lookup framework for Heathrow T3 capacity planning by route and departure time, and a Random Forest classifier (AUC 0.787) to predict booking completion. Key finding: purchase lead time is the strongest predictor of whether a customer completes a booking, actionable for targeted re-engagement campaigns. Findings were packaged into an executive presentation deck.

View on GitHub ↗ View Deck ↗
Online Sales Analytics
Python Python Jupyter Notebook Jupyter Notebook Tableau Tableau SQL SQL
2025

Online Sales Analytics

A CEO and CMO need data to decide where to expand and who to retain, but raw transactional data alone doesn't answer that. I cleaned and analyzed 540K+ retail transactions, reducing it to 397K reliable records, then built Tableau dashboards to surface strategic insights: revenue peaks in November and December signal an inventory planning opportunity, Netherlands, Ireland, and Germany lead international revenue, and a small group of repeat customers generates 3x more revenue during peak months, pointing to a clear ROI case for loyalty programs.

View on GitHub ↗
Loan Prediction Based on Customer Behavior
Python Python Jupyter Notebook Jupyter Notebook SQL SQL
2023

Loan Prediction Based on Customer Behavior

Manual loan approval processes are slow and inconsistent, and without a data-driven risk framework, lenders struggle to catch high-risk applicants early. As part of a 7-person team in Rakamin Academy's bootcamp, I contributed to building a loan default prediction model using the 5C credit analysis framework. Key findings: low-income customers carry a 15% default rate compared to 12% for high-income, single millennials show a 12.5% default risk, and renters default at a higher rate than homeowners — all actionable for credit scoring policy and automated approval systems.

View on GitHub ↗ View Deck ↗