nameAbdan Sakura
roleData Analyst
statusready_to_hire

Turning raw rows
into decisions.

Fresh graduate in Informatics Engineering, focused on data analysis, visualization & BI. Based in Depok, Indonesia.

Abdan Sakura
3.66
GPA / 4.00
4
Featured projects
3.9K+
Rows analyzed
6
Tools in the kit
01 — About

I'm a fresh graduate at Universitas Muhammadiyah Prof. Dr. Hamka with a strong interest in data, especially data analysis and data science. I enjoy working with data to find patterns and turn information into meaningful insight that supports decision-making. To me, data is a story waiting to be uncovered through the right approach.

My interest in mathematics and statistics helps me understand data more deeply, both in computational logic and result interpretation. I keep developing my skills across different tools and techniques, hoping to contribute data-driven solutions that create real impact.

EducationS1 Informatics Engineering
Univ. Muhammadiyah Prof. Dr. Hamka
Timeline2022 – 2026
FocusData Visualization & BI
LocationDepok, Indonesia
Org. roleSecretary, Social & Community Empowerment Div. — PK IMM FTII UHAMKA
02 — Skills & tools

What I work with

Data skills

Data Analysis
Data Visualization
Data Cleaning
Statistics
Exploratory Data Analysis

Tools & languages

Python — Pandas, NumPy, Matplotlib
SQL — SELECT, JOIN, GROUP BY
Tableau
Google Looker Studio
Microsoft Excel
Google Sheets
SPSS

Working style

Problem Solving
Fast Learner
Time Management
Attention to Detail
03 — Selected work

Projects

01
Google Looker Studio

Evaluating Subscription Program Effectiveness

A retail company's subscription program had never been evaluated with data, despite being central to its retention strategy — built from 3,900 transactions.

Key findings

  • Adoption is low and skewed by gender — 40% male vs. 0% female subscribed
  • Subscription doesn't lift spending or cut discount reliance ($59.5 vs. $59.9)
  • Revenue concentrates in customers aged 46+, in Clothing & Accessories

Recommendations

  • Redesign the program to be gender-inclusive
  • Shift from blanket discounts to loyalty-driven incentives
  • Prioritize retention campaigns for the 46+ segment
  • Explore demand-building for Outerwear
View project →
02
Tableau

Netflix TV Show Catalog Analysis

Netflix needs to understand its TV show catalog composition and performance to guide content acquisition, localization, and rating decisions.

Key findings

  • Dramas dominate the catalog — 1,600 titles, far ahead of other genres
  • The US leads production (~800 titles); international markets remain underdeveloped
  • TV-MA is the most common rating (1,145 titles); family content is minimal

Recommendations

  • Diversify ratings by investing in TV-G / TV-Y7 content
  • Expand non-US acquisition — Korea, Spain, France
  • Focus investment on high-performing genres while maintaining niche ones
View project →
03
Python — Pandas / Scikit-learn

Movie Recommendation System (Collaborative Filtering)

An item-based collaborative filtering engine on the MovieLens dataset, recommending movies by shared audience behavior rather than genre.

Users and movies with too few ratings are filtered out to keep the data manageable, then a sparse user-item matrix is built from what remains. Item-item cosine similarity measures how closely movies relate based on shared rating patterns, ranking the most similar titles for any query — searching "Toy Story" surfaces Toy Story 2, Forrest Gump, and Back to the Future, purely from shared audience taste.

View project →
04
Python — Decision Tree · Web Deployment

Mindscroll — Classifying TikTok Scrolling Behavior with Decision Tree

My undergraduate thesis classifies TikTok users into 3 mindless-scrolling categories (Low, Moderate, High) based on two variables: Scroll Behavior (frequency, duration, dwell-time, overuse) and Mindless Scrolling (entry point, normative dissociation, exit point), using survey data from 359 TikTok users aged 18+. After Data Selection, Preprocessing, Transformation, and Categorization stages, a Decision Tree model was tested across 3 scenarios — the best-performing model combined both variables (X+Y), reaching 81.48% accuracy and a 0.8875 macro AUC-ROC, with perfect recall (1.00) for the Low category. The model was then deployed into a website called "Mindscroll" so users can get their classification results directly.

✓ Copyright-registered (HKI) — Computer Program "Mindscroll"

View project →
04 — Education

Academic background

2022 — 2026

S1 Informatics Engineering — GPA 3.66

Universitas Muhammadiyah Prof. Dr. Hamka
  • Worked on data analysis projects using Python and SQL
  • Built data visualization dashboards in Tableau and Google Looker Studio
  • Studied fundamental machine learning concepts and applied them to datasets
05 — Experience

Where I've worked

Professional Experience

Mar — Jun 2026

Admin Data Entry

PT Satria Halal Indonesia
  • Input and submitted business-owner data into a web-based system
  • Verified document completeness — identity data, business registration (NIB), and supporting files
  • Validated submission data and flagged inconsistencies to support administrative review
Sep — Dec 2024

Digital Forensics Analyst — MSIB Batch 7

PT Analis Forensik Digital
  • Ran cyber-attack simulations and analyzed activity logs to identify potential threats
  • Acquired, identified, and analyzed digital evidence
  • Wrote data-driven investigation reports documenting findings and analysis

Organizational Experience

2024 — 2025

Secretary, Social & Community Empowerment Division

PK IMM FTII UHAMKA
  • Managed division administration and documentation
  • Contributed to planning and execution of work programs
  • Prepared activity reports and evaluated program outcomes
Kegiatan Pelita Aksara bersama PK IMM FTII UHAMKA

Let's talk about your data.

abdansakurati@gmail.com linkedin.com/in/abdan-sakura Depok, Indonesia Download CV