Skip to content
View Praruj's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report Praruj

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Praruj/README.md

English · 日本語

Praruj Thapa

AI/ML engineer and data scientist at Fukushima Galilei in Osaka. I build systems that turn noisy data into decisions people can trust; my current one flags equipment failures days before they happen.

All models are wrong, but some are useful. — George Box

Selected work

Writing

Python · PyTorch · scikit-learn · computational psychiatry · time series · search & recommendation · computer vision · RAG
Kyoto University of Advanced Science, B.E. 2025 · AWS Certified AI Practitioner

prarujthapa.com.np · LinkedIn · thapa.praruj@gmail.com


タパ プラルズ

大阪のフクシマガリレイ株式会社で働くAI・機械学習エンジニア/データサイエンティスト。ノイズの多いデータを、信頼できる判断へと変えるシステムを作っています。

  • 所属: フクシマガリレイ株式会社(時系列異常検知システムを開発。設備の故障兆候を3〜4日前に検知)
  • 専門: 時系列・センサーデータの機械学習、検索・レコメンド、Vision Transformer、モンテカルロシミュレーション、RAG・ローカルLLM
  • 学歴: 京都先端科学大学 工学部 機械電気システム工学科(2025年卒)

すべてのモデルは間違っているが、中には役に立つものもある。 — ジョージ・ボックス

主な取り組み

prarujthapa.com.np · LinkedIn · thapa.praruj@gmail.com

Pinned Loading

  1. stress-detection stress-detection Public

    Stress detection from wearable sensors (WESAD). CNN-LSTM + hybrid model, 92.7% on people the model never saw (leave-one-subject-out).

    Jupyter Notebook

  2. industrial-edge-anomaly industrial-edge-anomaly Public

    Industrial anomaly detection on the edge: PatchCore visual inspection (VisA) and vibration monitoring (CWRU), trained on healthy data only, int8 ONNX on one CPU thread.

    Python

  3. vision-transformer-projects vision-transformer-projects Public

    End-to-end CV projects: DETR hardhat detection + ViT knowledge distillation

    Jupyter Notebook

  4. vit-biomedical-finetuning vit-biomedical-finetuning Public

    Fine-tuning google/vit-base-patch16-224 (Vision Transformer) to classify peripheral blood cell microscopy images into 8 classes, using the Hugging Face transformers Trainer API. Trained on a free C…

    Jupyter Notebook

  5. worldcup_prediction_2026 worldcup_prediction_2026 Public

    World Cup 2026 prediction engine: Elo ratings + 100,000-run Monte Carlo simulation, Brier score 0.178.

    Python

  6. bilingual-product-search bilingual-product-search Public

    Japanese/English product search (BM25, multilingual E5, hybrid, reranking) on Amazon ESCI + purchase recommendation, with a FastAPI service

    Python