- Lowongan
- Panduan KarirPengembangan KarirWawancara Orang/PerusahaanBusiness ExcellenceDari CakeResumeBacaan Unggulan
- Untuk Perusahaan
- Download aplikasi CakeResume
CakeResume Talent Search
Paket Perekrutan Paling Mudah dan Efektif, Pilihan Ratusan Perusahaan
Cari lebih dari 800 ribu CV dan ambil aksi menghubungi pelamar kerja untuk rekrutmen yang lebih efektif. Pilihan ratusan perusahaan.
- Lihat semua hasil pencarian
- Tanpa batas harian untuk memulai pesan baru
- CV dapat diakses oleh perusahaan berbayar
- Lihat email pengguna & nomor telepon
Definition of Reputation Credits
Core Competencies
Hard Skills
- Python / Unix / Shell Script / SQL
- ML / DL / NLP
- Git / Docker
- Anaconda / Poetry
- GCP / AWS
- Retrieval-Augmented Generation (RAG)
- Vector Database (Qdrant, MongoDB Atlas)
- FastAPI
Soft Skills
- Leadership and mentorship
- Strong problem-solving ability
- Strategic Planning and Execution
- Self-learner
- Code Review and Quality Assurance
- Highly stress resistant
- Experience working in a startup and be part of an agile team
- Project management
- Mandarin (native) /English (TOEIC 840)
Familier Libraries
- Tensorflow / Keras
- Scikit-learn
- HuggingFace
- Pandas / Numpy
- Langchain
Uto.ai, Team Lead | Oct. 2023 - Present
At Uto.ai, I led the development of a multilingual, LLM-based chatbot that improved service delivery and user interaction across various countries. I managed a team of five engineers, focusing on complex, cross-functional communications in a fast-paced environment.
Leadership and Innovation
Spearheaded the AI team, focusing on talent acquisition, training, and mentorship. Implemented effective strategies for team growth and skill enhancement.
Project Management
Collaborated with stakeholders to align project goals with business objectives, ensuring cross-functional collaboration for project advancement.
Technical Expertise
Enhanced chatbot interactions by integrating vector databases with Retrieval-Augmented Generation (RAG), creating a customizable knowledge base, enhancing character interaction and user experience. Enabled support for multiple languages, catering to a diverse user base and showcasing the project's international reach.
Development of Long-Term Memory Systems
Utilized MongoDB to build a long-term memory system for the chatbot, enabling the retention and contextual use of information across interactions.
API Development
Led the development of RESTful APIs, ensuring seamless integration and communication between the chatbot and external services.
Quality Assurance and Best Practices
Conducted code reviews to maintain high standards of code quality, adherence to best practices, and to foster a culture of continuous learning and improvement within the team.
Playsee, AI Engineer, Mar. 2023 ~ Oct. 2023
At Playsee, I developed a conversational recommendation system that analyzed data from over 10 million users across multiple platforms. I designed text and video content recommendations tailored to international markets, thereby enhancing global user engagement.
Data Analysis at Scale
- Conducted in-depth analysis of extensive social media data, identifying user patterns and trends to inform the development of a conversational recommendation system.
Personalized Recommendations
- Leveraged the prowess of large language models (LLM), including ChatGPT, to guide users in articulating their preferences, enabling precise content recommendations.
- Collaborated directly with OpenAI engineers to stay at the forefront of LLM information and technology through bi-weekly webinar meetings.
- Analyzed user behavior with precision to deliver tailored recommendations, such as posts, reels and local stores.
- Formulated targeted advertising recommendations for commercial accounts in contexts that maximize their impact.
Vector Database Optimization
- Performed exhaustive benchmarking of multiple vector databases, discerning and implementing best practices to elevate system performance.
- Devised meticulous database schemas, finely tuned to achieve peak performance.
Sentence Embedding Service
- Evaluated a spectrum of embedding models, aligning selections with project requirements and effectively balancing performance and cost.
- Implemented GPU acceleration for expedited embedding computations while prudently managing budget constraints.
- Successfully orchestrated the deployment of the sentence embedding service onto Google Cloud Platform (GCP).
Emotibot, NLP Data Scientist, Sep. 2021 ~ Feb. 2023
As an NLP Data Scientist at Emotibot, I dedicated myself to advancing the field of Natural Language Processing (NLP) through research, innovation, and practical application. My role encompassed a wide range of responsibilities aimed at optimizing NLP downstream tasks, including Sentiment Analysis, Name Entity Recognition (NER), Relation Extraction, Key Information Extraction, and more. Here's a breakdown of my contributions:
Name Entity Recognition System
- Pioneered the development of a re-trainable platform for NER models, enabling fine-tuning with customized datasets.
- Successfully fine-tuned the baseline NER model, achieving an outstanding F1 score of up to 90%.
- Applied OCR results to extract critical information from receipts and tickets, enhancing data extraction capabilities.
Sentence Embedding
- Conceived and executed experiments utilizing pre-trained models to generate sentence embeddings, enhancing the performance of various NLP downstream tasks.
- Created an API server capable of receiving sentences and returning embeddings, streamlining the integration of pre-trained models across different modules, and improving operational efficiency.
- Implemented a groundbreaking document encoder that leveraged coordinates and sentence embeddings from bounding boxes as features.
- Incorporated BERT models to extract relationships between different tokens within documents, augmenting the understanding of textual content with spatial information.
- Established a dedicated sentence embedding API server to support seamless integration and utilization of these enhanced document understanding capabilities.
Chungyo Group, Machine Learning Engineer, Nov. 2020 ~ Apr. 2021
Manage a project of a real-time forecasting system for online games. Analyzed the data using Python libraries such as Pandas, Scikit-Learn, TensorFlow to extract features to analysis user-behavior and built statistical models in Python based on ML algorithms like SVM, Logistic Regression, Neural Networks.
Universal Scientific Industrial, Machine Learning Engineer, Oct. 2017 ~ Jun. 2020
Implementation of Object Detection and Face Recognition on embedding systems. These functions help users to classify photos automatically from the background with only a few hardware resources.
Education
National Cheng Kung University, M.Sc. in Institute of Computer and Communication Engineering, 2014 ~ 2016
National Central University, B.Cs. in Communication Engineering, 2009 ~ 2014
Self-learning
Natural Language Processing
Recommendation System
TensorFlow Developer Certificate, 2021/04
Testing the ability to use TensorFlow to build deep learning models for a range of tasks such as regression, computer vision, natural language processing, and time series forecasting.
Coursera Certifications (Online Courses)
Shopee Code League 2020
Shopee Code League - Product Detection - Kaggle
Use EfficientNet to detect products in images. Reach the top 30%.
Shopee Code League - Logistics - Kaggle
Cleaning data. Flagged out the late deliveries and penalties are imposed on the providers to ensure they perform their utmost.
Core Competencies
Hard Skills
- Python / Unix / Shell Script / SQL
- ML / DL / NLP
- Git / Docker
- Anaconda / Poetry
- GCP / AWS
- Retrieval-Augmented Generation (RAG)
- Vector Database (Qdrant, MongoDB Atlas)
- FastAPI
Soft Skills
- Leadership and mentorship
- Strong problem-solving ability
- Strategic Planning and Execution
- Self-learner
- Code Review and Quality Assurance
- Highly stress resistant
- Experience working in a startup and be part of an agile team
- Project management
- Mandarin (native) /English (TOEIC 840)
Familier Libraries
- Tensorflow / Keras
- Scikit-learn
- HuggingFace
- Pandas / Numpy
- Langchain
Uto.ai, Team Lead | Oct. 2023 - Present
At Uto.ai, I led the development of a multilingual, LLM-based chatbot that improved service delivery and user interaction across various countries. I managed a team of five engineers, focusing on complex, cross-functional communications in a fast-paced environment.
Leadership and Innovation
Spearheaded the AI team, focusing on talent acquisition, training, and mentorship. Implemented effective strategies for team growth and skill enhancement.
Project Management
Collaborated with stakeholders to align project goals with business objectives, ensuring cross-functional collaboration for project advancement.
Technical Expertise
Enhanced chatbot interactions by integrating vector databases with Retrieval-Augmented Generation (RAG), creating a customizable knowledge base, enhancing character interaction and user experience. Enabled support for multiple languages, catering to a diverse user base and showcasing the project's international reach.
Development of Long-Term Memory Systems
Utilized MongoDB to build a long-term memory system for the chatbot, enabling the retention and contextual use of information across interactions.
API Development
Led the development of RESTful APIs, ensuring seamless integration and communication between the chatbot and external services.
Quality Assurance and Best Practices
Conducted code reviews to maintain high standards of code quality, adherence to best practices, and to foster a culture of continuous learning and improvement within the team.
Playsee, AI Engineer, Mar. 2023 ~ Oct. 2023
At Playsee, I developed a conversational recommendation system that analyzed data from over 10 million users across multiple platforms. I designed text and video content recommendations tailored to international markets, thereby enhancing global user engagement.
Data Analysis at Scale
- Conducted in-depth analysis of extensive social media data, identifying user patterns and trends to inform the development of a conversational recommendation system.
Personalized Recommendations
- Leveraged the prowess of large language models (LLM), including ChatGPT, to guide users in articulating their preferences, enabling precise content recommendations.
- Collaborated directly with OpenAI engineers to stay at the forefront of LLM information and technology through bi-weekly webinar meetings.
- Analyzed user behavior with precision to deliver tailored recommendations, such as posts, reels and local stores.
- Formulated targeted advertising recommendations for commercial accounts in contexts that maximize their impact.
Vector Database Optimization
- Performed exhaustive benchmarking of multiple vector databases, discerning and implementing best practices to elevate system performance.
- Devised meticulous database schemas, finely tuned to achieve peak performance.
Sentence Embedding Service
- Evaluated a spectrum of embedding models, aligning selections with project requirements and effectively balancing performance and cost.
- Implemented GPU acceleration for expedited embedding computations while prudently managing budget constraints.
- Successfully orchestrated the deployment of the sentence embedding service onto Google Cloud Platform (GCP).
Emotibot, NLP Data Scientist, Sep. 2021 ~ Feb. 2023
As an NLP Data Scientist at Emotibot, I dedicated myself to advancing the field of Natural Language Processing (NLP) through research, innovation, and practical application. My role encompassed a wide range of responsibilities aimed at optimizing NLP downstream tasks, including Sentiment Analysis, Name Entity Recognition (NER), Relation Extraction, Key Information Extraction, and more. Here's a breakdown of my contributions:
Name Entity Recognition System
- Pioneered the development of a re-trainable platform for NER models, enabling fine-tuning with customized datasets.
- Successfully fine-tuned the baseline NER model, achieving an outstanding F1 score of up to 90%.
- Applied OCR results to extract critical information from receipts and tickets, enhancing data extraction capabilities.
Sentence Embedding
- Conceived and executed experiments utilizing pre-trained models to generate sentence embeddings, enhancing the performance of various NLP downstream tasks.
- Created an API server capable of receiving sentences and returning embeddings, streamlining the integration of pre-trained models across different modules, and improving operational efficiency.
- Implemented a groundbreaking document encoder that leveraged coordinates and sentence embeddings from bounding boxes as features.
- Incorporated BERT models to extract relationships between different tokens within documents, augmenting the understanding of textual content with spatial information.
- Established a dedicated sentence embedding API server to support seamless integration and utilization of these enhanced document understanding capabilities.
Chungyo Group, Machine Learning Engineer, Nov. 2020 ~ Apr. 2021
Manage a project of a real-time forecasting system for online games. Analyzed the data using Python libraries such as Pandas, Scikit-Learn, TensorFlow to extract features to analysis user-behavior and built statistical models in Python based on ML algorithms like SVM, Logistic Regression, Neural Networks.
Universal Scientific Industrial, Machine Learning Engineer, Oct. 2017 ~ Jun. 2020
Implementation of Object Detection and Face Recognition on embedding systems. These functions help users to classify photos automatically from the background with only a few hardware resources.
Education
National Cheng Kung University, M.Sc. in Institute of Computer and Communication Engineering, 2014 ~ 2016
National Central University, B.Cs. in Communication Engineering, 2009 ~ 2014
Self-learning
Natural Language Processing
Recommendation System
TensorFlow Developer Certificate, 2021/04
Testing the ability to use TensorFlow to build deep learning models for a range of tasks such as regression, computer vision, natural language processing, and time series forecasting.
Coursera Certifications (Online Courses)
Shopee Code League 2020
Shopee Code League - Product Detection - Kaggle
Use EfficientNet to detect products in images. Reach the top 30%.
Shopee Code League - Logistics - Kaggle
Cleaning data. Flagged out the late deliveries and penalties are imposed on the providers to ensure they perform their utmost.