Indonesia

Data Scientist

AI/ML Engineering

Bachelor's degree

Ruben Stefanus

Senior Data Scientist

Communication

Problem-solving

Critical-thinking

Analytical skills

AI/ML

Cloud computing

Pandas

Python

PyTorch

Docker

Jupyter Notebook

NLTK

Numpy

TensorFlow

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build, enhance and maintain Engine to extract itemized billing from Hospital Invoices (OCR + classifier model)
Problem: Manual process in registering itemized billing from invoice
𝐔𝐬𝐞𝐫: Insurance -- Third-Party Administrator (TPA) business
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Automate itemized billing process
• Reduced the time required to input itemized billing items


𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build, enhance and maintain Medicine Recommendation Engine for Pharmacy Delivery business
𝐔𝐬𝐞𝐫: Halodoc users that buy medicine in "Toko Kesehatan" feature
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Increased basket size of order
• Increased revenue

𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐢𝐛𝐢𝐥𝐢𝐭𝐢𝐞𝐬:
• Work with multiple and complex data sources from financial institutions (e.g. Bank, P2P Lending, etc)
• Processing, cleansing, and validating the integrity of data to be used for analysis and model development
• Collaborate with engineering and product team in data products development


𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build, enhance and maintain transaction classifier product (classify bank transaction data)
𝐄𝐱𝐚𝐦𝐩𝐥𝐞: KARTU DEBIT STARBUCKS KEMANG, 5260XXXXXXXX
𝐑𝐞𝐬𝐮𝐥𝐭: EXPENSE - DEBIT - STARBUCKS - KEMANG
𝐔𝐬𝐞𝐫: Personal Finance Management (PFM) companies
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Understanding a person's behavior based on bank transaction data
• Generate summary of personal cash flow
 
𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build, enhance and maintain bank OCR product (OCR for bank statement pdf file)
𝐔𝐬𝐞𝐫: Lending companies
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Reduced manual effort in data extraction process

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Research and Build income prediction product

1-on-1 mentoring, give private consultation about real case in daily data practitioner job

• Actively do research project and publish it
• Speaker at Melek For Member (MFM) Donation Event

A research community that aims to implement and develop Artificial Intelligence in Indonesia
• Responsible for managing, running and in charge of community agenda
• Actively do a research project and publish article

[ 𝐁𝐫𝐢𝐧𝐠𝐢𝐧𝐠 𝐫𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐚𝐧𝐝 𝐢𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐨𝐧 𝐢𝐧𝐭𝐨 𝐭𝐡𝐞 𝐫𝐞𝐚𝐥 𝐰𝐨𝐫𝐥𝐝 ]

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build product-name matching model between internal and competitor products
𝐔𝐬𝐞𝐫: pricing manager, catalog team
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Market Price Visibility
• Find out products that are not available internally

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build demand prediction model for several SKU products
𝐔𝐬𝐞𝐫: buyer team
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Successfully assisted the buyer team in determining the number of products to restock
• Reduced out-of-stock event for certain SKUs
• Reduced unfulfilled demand

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build OCR model for Indonesian ID Card that recognizes NIK and Name to help user verification process
𝐔𝐬𝐞𝐫: user verification team
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Reduced manual work that required by user verification team
• Reduced user verification time

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Build "Kubeflow Watcher" that integrates Kubeflow, internal chatbot, and company chat app
𝐔𝐬𝐞𝐫: Data Scientist, Machine Learning Engineer
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬:
• Provides visibility for KubeFlow Pipeline run status
• Alert system for failed pipeline

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Research and build an NLP model that has a possibility to embed with our business (NER for Warung Order)

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Implementation of TIBCO Data Science platform for Indonesian telecom company
𝐔𝐬𝐞𝐫: Data Scientists
𝐃𝐞𝐭𝐚𝐢𝐥𝐬:
• End-to-end implementation of TIBCO DS platform and train users how to use the platform
• Integration TIBCO DS platform with SQL databases (i.e. Greenplum, PostgreSQL, Teradata, Oracle) and Big Data frameworks (i.e. Hadoop, Cloudera, Spark, Hive)
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬: 
• Seamless migration to new platform
• Reduced data science project development time

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: Refactor existing ETL pipelines and new data sources integration for Indonesian telecom company
𝐔𝐬𝐞𝐫: Business Intelligence Analyst
𝐃𝐞𝐭𝐚𝐢𝐥𝐬:
• Refactoring existing ETL pipelines
• Data sources integration (i.e. Teradata, Greenplum, Oracle, HDFS)
• QA Testing for result table of ETL pipelines
𝐀𝐜𝐡𝐢𝐞𝐯𝐞𝐦𝐞𝐧𝐭𝐬: 
• Reduced ETL pipeline complexity
• Reduced ETL processing time

Guide and assist the difficulties faced by participants

•  Content of workshop is about learn Hadoop architecture, installation and configuration Hadoop on single cluster 
• Introduction about YARN, HDFS, RDD then create MapReduce job and Spark job using Python

• Teach about the AVR Microcontroller
• Arduino Programming
• Various kinds of sensors and their implementation in the real world
• How to build a robot
• Deliver and explain about the concept, implementation, and guide students

Ruben Stefanus🇮🇩

About Ruben Stefanus

Expertise

Sessions with Ruben Stefanus

What mentees say