Natapon Pantuwong

Python Developer and Machine Learning Engineer

Bangkok, Thailand

Summary

I have 10 years of experience in Python. Currently, my main development is NLP and machine learning service API for clients. I am an enthusiastic and responsible person. I also have experience as a remote developer. Please consider my profile and background.

Languages:

English, Japanese, Thai

Favorite Python Packages:

Scikit-learn, numpy, scipy, flask, grpcio, opencv, tensorflow, pytorch

Experience

My current interests are including topics related to machine learning. One of them is NLP for Thai language. There are several applications and algorithms are in focused now such as Thai sentences classification, Name Entity Recognition for Thai language, and Machine translation for Thai language. These NLP techniques are focused to develop a Chatbot for Thai language. Apart from the NLP, I am also interested in machine learning for vision. The objective of current researches is the development of algorithms to support movie visual effect creation in the movie industries. The examples of topics are template matching for object tracking using GPU, object tracking using CNN on GPU, statistical calculation for object tracking across frames of the movie shot. I am also interested in GPU-based algorithm design to improve the time performance of the machine learning model. Apart from the above topics, I am also interested in investigating new technologies to support automated system in the software development life cycle such as CI/CD using Jenkins, Docker, ELK stacks, and so on.

Here are some of my previous projects.

• Sea Turtle Identification from Photos: In this project, I would like to identify the existing of sea turtle which could be found naturally using Photo ID. The recognition model is created from the feature points extracted from sea turtle photos. With Tensorflow object recognition API, we can achieve a good identification result.

• Sentence Classification for Thai Language: I would like to create an automatic question/answer system as a Chatbot to perform a help desk task. Therefore, I create a classification model to classify the sentence in Thai language. This is completed using Python, Scikit-learn, and BERT
• Name Entity Recognition for Thai Language: In a Chatbot, I need to recognize several entities in the sentence to understand the context and detail of the sentence. I extend the existing Duckling framework from Facebook to Thai language. Currently, my implementation can recognize numeral, currency, and time domain of Thai language. This is completed using Haskell.

• Machine Translation API for Thai Chatbot: With emerging of Neural Machine Translation services provide by Google (Google AutoML) and Microsoft (Azure Custom Translation), we can achieve machine translation by using the model trained by such services. Because of pricing condition, I choose to use Microsoft one. I then create a REST API or GRPC API using Python for the Chatbot client

Skills

Artificial Intelligence (AI), Backend Development, Data Science, Flask, Git, JavaScript, Linux, Machine Learning, NumPy, Pandas, SQL, SciPy, TensorFlow, jQuery