Vinson Ciawandy is a Data Scientist from Indonesia. He begun his Data Science journey in 2017, when one of the biggest E-commerce company came to the campus to gave workshop about Data Analytics. Now, he works as a Data Scientist in the same company who introduced him to Data Science.
Work Experience
-
Data Scientist, Tokopedia
Fulltime, February 2021 - Present -
Data Scientist, PT Mandala Multifinance
Contract, June 2020 - January 2021
Take part to deliver digital transformation to Mandala Multifinance:-
Created Deep Learning model using Tensorflow with distributed GPU training strategy and automatic model evaluation with Champion/Challenger approach for credit scoring.
-
Created dashboard to monitor model performance using Dash
-
Created dashboard to generate report using Streamlit
-
Created scoring prediction services that connected to Kafka services
-
Developed Anomaly Detection worflow
-
-
Data Analyst, Purple Analytics
Internship, June 2019 - August 2020
With a team of 2, we created :-
Recommendation system using KNN, Matrix Factorization and Deep Learning
-
Web mockup for showing recommendation system result using Django and successfully deploy it to local server and Heroku.
-
Heatmaps generation for people detection analysis
-
-
Lab Assistant of Statistics, ITB
Part time, September 2018 - December 2019
Help students to apply statistical methods such as :-
Hypothesis Testing
-
Linear Regression
-
Time Series Analysis
-
Spatial Analysis
-
Statistical Process Control
in R programming language for Data Analysis and Basic Statistics courses.
-
-
Data Scientist, PT Telkom Indonesia
Internship, May 2018 - July 2019
Worked on a joint project between PT Telekomunikasi Indonesia with PT Jasa Marga to improve vehicle counting system on rest area.-
Created daily report using Tableau for daily monitoring
-
Identified the problem with the current system
-
Created machine learning model and demo
-
Education
-
B.Sc in Mathematics, Institut Teknologi Bandung
August 2016 - October 2020, GPA 3.66 (Cum Laude)
Undergrad Thesis : Human Motion Trajectory Prediction with Deep Learning
Most of the classes I took related to Statistics. I also took some class from Computer Science(Data Structure and Databases). Active participation in Data Science competitions and student organization. -
Santa Maria Pekanbaru Senior High School
2013 - 2016
Taking a part in Genta Honggoi Orchestra as the guitar player was the best decision I took during my highschool :)
Workshop given
-
Speaker at “Cashless Society, Sudah Siapkah Indonesia?" talks
KM ITB, November 2019
Became one of four speakers at KM ITB’s LOKAKARYA INOVASI II discussing cashless society in Indonesia. -
Instructor at “Turn Data into Insight”’ workshop.
LnPoint, November 2019
Created a 3 days workshop to taught the basics of Data Science. The material that I covered including : Data wrangling, Data visualization and Machine Learning. -
Instructor at “Gentle Introduction to Machine Learning : Supervised Learning”’ workshop.
LnPoint, March 2019
Created a 1 days workshop to taught the basics of supervised learning using R programming language.
Skill and Knowledge
- Data Science, Mathematics, Statistics, Computer Vision,Recommender System, Data Structures
- Python, R, SQL, C
- Pytorch, Tensorflow, Plotly-Dash, MLflow, Kafka
- Tableau, Apache Superset, Git
Competition Experience
-
2nd posititon at PETSROCHINA Paper and Exhibition Competition
APECX UGM, November 2019
In a team of 3, we wrote “Inferring Production Rates of a Rod Pumped Well Using Keras Neural Network and Random Forest” -
1st position at Data Analysis Competition
IFest Unpad, October 2019
In a team of 4, we created an analysis and dashboard about Indonesia’s education quality. We conclude that National Exam for high school students is ineffective to measure student’s competence and we debunk some common belief about education. -
2nd position at Data Mining Competition
Joints UGM, May 2019
Given data from Mamikos, our objective is to predict the right price to give rent a boarding house(kost). Things I do includes : Imputate missing value through Data Exploration(by using Alluvial plot, correlation and geographical fact), create and optimize Gradient Boosting Machine to predict the right price -
Finalist In-Country at EY NextWave Data Science Challenge
Ernst & Young, May 2019
Given trajectory record of mobile device in Atlanta City , we asked to predict whether the device will leave the city or not. We convert the historical data into supervised learning problem, done tons of feature engineering and feed it into Gradient Boosting Machine
I’ll just keep moving forward Eren Jäger
Hobby
- Writing this blog(New hobby)
- Hash run(But I don’t drink) and badminton
- Playing JRPG games (Huge fan of all the trails series)