Charin Polpanumas
Senior Applied Scientist at Amazon
Tokyo, Japan
Summary
🤩
Rockstar🎓
Top School「おはようウサギ🐰」Charin Polpanumas is a Senior Applied Scientist based in Tokyo with a decade of experience building AI-driven decision systems that translate causal inference and heterogeneous treatment effect estimation into millions in annual revenue. At Amazon he designs personalized customer experiences and simulates customer behavior with LLMs, bridging offline modeling and real-world feedback to keep decisions robust under uncertainty. Previously he led data science at Central Group, shipping Thai-language search, recommendation, and CRM systems, and contributed important Thai NLP datasets and tooling to open-source projects like Hugging Face datasets and PyThaiNLP. He prefers deep exploratory analysis and interpretable models, and emphasizes knowing when not to trust a model as much as improving it. Fluent in four languages and experienced across production stacks (SQL, Python, PyTorch, R), he combines applied research rigor with hands-on engineering to deploy high-impact, reliable ML products.
10 years of coding experience
8 years of employment as a software developer
Bachelor's Degree, Economics (Marcoeconomics), Bachelor's Degree, Economics (Marcoeconomics) at 一橋大学Master's Degree, Management (Marketing Research), 92.8%, Master's Degree, Management (Marketing Research), 92.8% at Peking University
English, Japanese, Chinese, Thai
Github Skills (15)
machine-learning10
nlp10
nlpjs10
python10
natural-language-processing10
text-processing10
datasets10
data-science9
data-engineering9
fastai9
machine-learning-models8
tensorflow7
pytorch7
pandas6
numpy6
Programming languages (7)
Github contributions (5)
PyThaiNLP/pythainlp
Jun 2018 - May 2022
Thai natural language processing in Python
Role in this project:
Back-end Developer
Back-end DeveloperContributions:4 reviews, 43 commits, 19 PRs in 3 years 11 months
Contributions summary:Charin primarily focused on enhancing the Thai natural language processing library by modifying the download location of a model and adding documentation for a specific module (ulmfit). Their contributions also involved updating dependencies, modifying tokenizer arguments, and incorporating updated pretrained model paths. Furthermore, the user was involved in updating and modifying some existing code in the core of the `pythainlp` library.
nlpword-segmentationpythonthai-nlp-librarylanguage-processing
huggingface/datasets
Dec 2020 - Feb 2021
🤗 The largest hub of ready-to-use datasets for ML models with fast, easy-to-use and efficient data manipulation tools
Role in this project:Data Scientist & ML Engineer
Data Scientist & ML EngineerContributions:5 reviews, 17 commits, 23 PRs in 2 months
Contributions summary:Charin contributed significantly to the addition of several datasets related to Thai language processing and machine learning, primarily for NLP tasks. Their contributions included the integration of datasets like `prachathai67k`, `wisesight_sentiment`, `generated_reviews_enth`, `scb_mt_enth_2020`, `thaisum`, `thainer`, `thaiqa_squad`, `lst20`, `wisesight1000`, `thai_toxicity_tweet`, `best2009`, `stereoset`, `wongnai_reviews`, and `iapp_wiki_qa_squad`. These additions involved the creation of dataset configurations, data loading logic, and feature definitions, demonstrating a focus on preparing data for machine learning models within the Hugging Face datasets library.
ml-modelstensorflownatural-language-processingmanipulationdata-science
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.