Graduated from NTHU DSPMT, with the double specialties of quantitative finance and computer science. Focusing on web development and data science.
cellphone: 0979164823 [email protected]
Shao Hsiang Chien Portfolio ShaoXiangChien
- Built MLOps Lab that achieves IaC, ML CICD, auto retrain trigger, deployments on ACI and AKS, and monitoring with Application Insight, for enterprise customers using Azure Machine Learning and Azure DevOps, and delivered the workshop in 3 different sessions
- Implemented PoC of multistop routing with Azure Maps for the logistics of our e-commerce client, as well as the demos of Azure Cognitive Search and Anomaly Detector
- Built the core functionality of the project, which enables people to team up as a startup group on the platform using React and Firebase
- Led a team of 5 developers and introduce Scrum into our team using Trello for better collaboration
- Web Development: HTML | CSS | JavaScript | React | Firebase | Django | RWD | MUI
- Data Science: Python | Tensorflow | Data Engineering | ETL Pipeline | SQL | R
- Azure: DevOps | CICD | Pipeline | Azure Machine Learning | Cognitive Service | App Service
- Tools: Git Version Control | GitHub | Linux | DevTools
Azure OpenAI Enabled FAQ Bot - Parsed the information from URLs with Custom Question Answering service and deployed the service as a chatbot with Azure App Service and Web App Bot. Enhanced the response of the bot with OpenAI Davinci Model and ChatGPT, and turned the user query and the question set into Text Embeddings to map the query to the questions in a semantic way. Created a UI with Streamilt. (03/2023)
Timeline Generator - Gather and process data of a given event through ETL Pipeline (with BeautifulSoup, Pandas, and RE), adopt Anomaly Detector to identify the important timestamps, and apply a cluster-based method alongside BERT Embeddings to extract the final incidents composing the timeline. (02/2023)
2019 - 2023
Apr 2018 - Jan 2020
Developed a platform for the audience of Netflix dramas to check out opinions about a certain drama from different sources, utilizing different data operation techniques, such as web crawling, data integration, and sentiment analysis.
Jan 2021 - Dec 2021
Performed data mining on the sales datasets of Pchome to find their pain point and research for the prospective customers as well as their buy-back cycles.
Jan 2021
Performed business analysis on a social enterprise, DiD HK, and proposed a strategy to solve their issues regarding their lack of social influence and reputation.
Find out the prospective consumers and their buy-back cycles through data mining. Formulate solutions with valuable insights from the analysis of data.
Collect and integrate public opinions about Netflix drama through web crawling, and provides sentiment analysis results of the comments. Presented with a website made with Streamlit and Firebase
A proof of concept for MLOps with Azure. Achieved Infrastructure As Code, Auto Retrain Trigger, and Machine Learning Development CICD.
A startup platform for people who want to start a business or are eager to join a startup team, providing matchmaking between teams and members, a sandbox to simulate the whole startup process alongside courses, and a platform for product exposures.
Graduated from NTHU DSPMT, with the double specialties of quantitative finance and computer science. Focusing on web development and data science.
cellphone: 0979164823 [email protected]
Shao Hsiang Chien Portfolio ShaoXiangChien
- Built MLOps Lab that achieves IaC, ML CICD, auto retrain trigger, deployments on ACI and AKS, and monitoring with Application Insight, for enterprise customers using Azure Machine Learning and Azure DevOps, and delivered the workshop in 3 different sessions
- Implemented PoC of multistop routing with Azure Maps for the logistics of our e-commerce client, as well as the demos of Azure Cognitive Search and Anomaly Detector
- Built the core functionality of the project, which enables people to team up as a startup group on the platform using React and Firebase
- Led a team of 5 developers and introduce Scrum into our team using Trello for better collaboration
- Web Development: HTML | CSS | JavaScript | React | Firebase | Django | RWD | MUI
- Data Science: Python | Tensorflow | Data Engineering | ETL Pipeline | SQL | R
- Azure: DevOps | CICD | Pipeline | Azure Machine Learning | Cognitive Service | App Service
- Tools: Git Version Control | GitHub | Linux | DevTools
Azure OpenAI Enabled FAQ Bot - Parsed the information from URLs with Custom Question Answering service and deployed the service as a chatbot with Azure App Service and Web App Bot. Enhanced the response of the bot with OpenAI Davinci Model and ChatGPT, and turned the user query and the question set into Text Embeddings to map the query to the questions in a semantic way. Created a UI with Streamilt. (03/2023)
Timeline Generator - Gather and process data of a given event through ETL Pipeline (with BeautifulSoup, Pandas, and RE), adopt Anomaly Detector to identify the important timestamps, and apply a cluster-based method alongside BERT Embeddings to extract the final incidents composing the timeline. (02/2023)
2019 - 2023
Apr 2018 - Jan 2020
Developed a platform for the audience of Netflix dramas to check out opinions about a certain drama from different sources, utilizing different data operation techniques, such as web crawling, data integration, and sentiment analysis.
Jan 2021 - Dec 2021
Performed data mining on the sales datasets of Pchome to find their pain point and research for the prospective customers as well as their buy-back cycles.
Jan 2021
Performed business analysis on a social enterprise, DiD HK, and proposed a strategy to solve their issues regarding their lack of social influence and reputation.
Find out the prospective consumers and their buy-back cycles through data mining. Formulate solutions with valuable insights from the analysis of data.
Collect and integrate public opinions about Netflix drama through web crawling, and provides sentiment analysis results of the comments. Presented with a website made with Streamlit and Firebase
A proof of concept for MLOps with Azure. Achieved Infrastructure As Code, Auto Retrain Trigger, and Machine Learning Development CICD.
A startup platform for people who want to start a business or are eager to join a startup team, providing matchmaking between teams and members, a sandbox to simulate the whole startup process alongside courses, and a platform for product exposures.