Dea Ramanda

Get to know me

About

I'm a data analyst who likes finding the story hiding inside a spreadsheet. I spend most of my time turning raw logs into dashboards people actually want to open — and I care more about the "why" behind a number than the number itself. Currently exploring business analytics.

📍 Based in

Jakarta, Indonesia

UTC +7 · WIB

🛠 Stack & tools

Python SQL Pandas NumPy Scikit-learn TensorFlow Streamlit Plotly Power BI PostgreSQL

🎨 Design & other

Figma Webflow Framer Illustrator

If we share these, say hi

📖 Reading (my Goodreads is below)

✍️ Writing on my blog

🎻 Classical music — DM me for recs

Selected work

Featured Projects

A cross-section of the problems I like solving — turning operational data, imagery, and user behavior into things people can use. Each one links out to the full case study including business context, the data journey, and what it changed.

Project 01

Exploratory Data Analysis · Demand Forecasting

Bike Rental, is it really growing?

An interactive analytics dashboard exploring how weather conditions, seasonal shifts, weekdays, and peak hours drive urban bike-sharing demand.

Business problem: Unpredictable commute spikes caused bike shortages at high-demand stations during morning/evening rush hours, while other stations remained overstocked, leading to lost rental revenue and high rebalancing costs.

Python Pandas Seaborn Streamlit

View Dashboard

Project 02

Data Storytelling · Operations Analytics

Can One Ship Break Global Trade?

A deep-dive into vessel and port logs to test whether a single canal blockage really disrupts shipping operations worldwide — or whether the bigger bottlenecks were hiding somewhere else entirely.

Business problem: Supply chains are planned around the assumption that shipping is predictable. Leadership needed evidence on what actually drives delays before investing in a fix.

Python Pandas Plotly

View Dashboard

Daily canal transits Mar 2021

Project 03

Business Intelligence · Data Visualization

Market Sales Performance Dashboard

An interactive BI dashboard built with Streamlit, Pandas, and Plotly to analyze sales performance, customer behavior, and product insights — turning transactional data into decisions that support revenue growth.

Business problem: Leadership lacked visibility into which products and customers actually drove growth, and relied on gut feel instead of transactional data.

Python Plotly Streamlit Pandas

Live Demo

$482K ▲ 18% QoQ

Sentinel-2 · AWEI_sh 2019 → 2026

Project 04

Remote Sensing · Geospatial Analytics

Coastal Land Change — North Jakarta

A spatial-temporal analysis of coastal land cover dynamics from 2019–2026 using Sentinel-2 imagery and Google Earth Engine, tracking land reclamation and coastal development over time.

Business problem: Coastal authorities lacked consistent, multi-year visibility into reclamation and land-cover change needed for flood-risk and permitting decisions.

Google Earth Engine Python Sentinel-2 Geemap

Live Demo

Project 05

Computer Vision · Transfer Learning

Reusable vs. Disposable Container Classifier

A thesis project applying transfer learning with MobileNetV2 to classify tumbler use, supporting environmental-awareness programs through AI-powered detection. Deployed publicly on Hugging Face.

Business problem: Sustainability programs had no automated way to verify reusable-container use at scale, limiting how well they could track and reward the right behavior.

Python TensorFlow/Keras MobileNetV2 HuggingFace

Live Demo

Project 06

Computer Vision · Deep Learning

Tabir — Real-Time Sign Language Detection

An AI-powered SIBI (Indonesian Sign Language) translation system built on a custom video dataset, designed to reduce everyday communication barriers for the deaf community.

Business problem: Deaf and hard-of-hearing individuals face real communication barriers without affordable, real-time translation tools.

Python Keras TensorFlow MediaPipe

Live Demo

Project 07

Machine Learning · Recommender System

Memorabilia — Book Recommendation System

A personalized book recommendation engine using Neural Collaborative Filtering, built to lift reading engagement and literacy as part of the Memorabilia mobile app.

Business problem: Reading engagement was low, and generic bestseller lists weren't bringing readers back to the app.

Python TensorFlow/Keras Pandas Flask

Live Demo

Continuous learning

What I've Learned

My journey

Experience

Jul 2025 – Oct 2025

Workflow Administrator

Sinarmas Distribusi Nusantara · Contract

Managed approval workflows for high-value transactions, new customer onboarding, and credit-risk cases across nationwide FMCG distribution operations, and analyzed AR records to support credit decisions.

Feb 2025 – Jul 2025

Frontend & Backend Developer Trainee

DBS Foundation × Dicoding Coding Camp

Built responsive single-page applications with JavaScript, Webpack, and Tailwind CSS following modular development practices.

Credential

Feb 2024 – Jun 2024

Machine Learning Cohort

Bangkit Academy — led by Google, GoTo, Traveloka

Built end-to-end ML solutions spanning preprocessing, feature engineering, training, and deployment with TensorFlow and Scikit-learn. Graduated with Distinction (Top 10%).

Credential

2021 – 2025

B.Sc. in Informatics

Universitas Bina Sarana Informatika — GPA 3.68/4.00

Also holds a Program Analyst Certification (BNSP, 2024–2027) and a MikroTik Certified Network Associate certification (2025–2028).

Dea Ramanda

About

Dea Ramanda

Featured Projects

Bike Rental, is it really growing?

Can One Ship Break Global Trade?

Market Sales Performance Dashboard

Coastal Land Change — North Jakarta

Reusable vs. Disposable Container Classifier

Tabir — Real-Time Sign Language Detection

Memorabilia — Book Recommendation System

What I've Learned

Experience

Workflow Administrator

Frontend & Backend Developer Trainee

Machine Learning Cohort

B.Sc. in Informatics

Contact