Use0hzx0oye3ersbch4l

洪瑋辰 Esther Hung

Machine Learning Engineer | Python Engineer | Cloud Solutions Architect for AWS

Studied Machine Learning in an AI Lab


[email protected] 
(+886) 937294582


Skills


  • Python
    • Web crawler (selenium and requests)
    • Organize web crawler results (beautifulsoup)
    • Data ETL
    • Simple web (flask)
    • Simple LineBot for setting notifications (Django and linebot)
    • Create numerous AWS CloudWatch alarms (boto3)
    • Auto run numerous Weka models and organize results to tables
    • Auto send data, get results, and organize results to tables for Pse-in-one 2.0
    • Switch amino acid to representing values based on AAindex rules
    • Rasa actions 
    • Create independent environment (venv)
    • Test chatbot NLU model performance (psycopg2, pandas, requests)
    • Change chatbot old format to new format
    • Remote reboot host while ssh does not work (SimpleHTTPServer)
  • Machine Learning
    • Do feature encoding for data with related informations
    • Build prediction models with LibSVM, Weka, MLP, CNN, RNN, LSTM
    • Keras and Tensorflow for building models
    • Do independent tests, cross validations, and case study for models
    • Evaluate models with ACC, AUC, MCC, F1, Sensitivity (Sn), and Specificity (Sp)
    • Modify number of iterations (epochs) and batch size for optimizing models
    • Set early stopping to avoid overtraining
    • Modify pooling size, learning rate, kernel or filter size, and dropout rate for optimizing models
    • Modify activation functions, optimizers, and loss functions for optimizing models
    • Modify number of hidden layers and number of activation units for optimizing models
    • Modify LibSVM code for adding additional evaluation metrics (C)
    • Scikit-learn for drawing ROC curve and calculate
    • Rasa framework for chatbot
  • AWS
    • IAM (create users for customer's consulters, create roles for aws services, and create policies for users or roles based on least privilege)
    • VPC (create subnets, set route table, set NAT gateway, create endpoint, and create Client VPN)
    • EC2 (write user data, set security group, add hosts, install and set cloudwatch agent, set auto start and stop, change a hostname, change a password, add to an AD, create images, set lifecycle manager, set auto backup for Microsoft SQL Server database, and set related setting for connection)
    • S3 (create buckets, set bucket policies, and set lifecycle managements)
    • ELB (set listeners, edit listener rules, target groups, and set security group)
    • ECS (set task definitions, create clusters, set services, connect load balancers, and set auto scaling)
    • CloudWatch (create dashboards and set alarms)
    • QuickSight (design simple dashbroads, create datasources, and design datasets)
    • Lambda (set Client VPN whitelist, auto refresh QuickSight, and create numerous CloudWatch alarms)
    • Systems Manager (set auto start and stop EC2, set auto reboot EC2, and reset EC2 password for forgetting password)
    • WorkSpaces (create workspaces, create images, create bundlers, and set IP access controls)
    • SDK (AWSCLI, AWSPowerShell, and boto3)
    • WAF (basic setting and modify AWS-managed rules)
    • Route 53 (basic setting and manage forwarding rules)
    • CloudFront (basic setting and set security header)
  • Host Management
    • Prometheus and Grafana for monitor (CPU, Memory, GPU, Disk Write/Read, Disk available bytes, Service status)
    • Node-RED for scheduler
    • Nginx for reverse proxy
    • Docker for runtime environment and database
    • Openvpn server management
    • Account management for hosts and applications
    • Database management for PostgreSQL and MongoDB
    • Write some scripts to backup system and application logs
    • Update OS to latest version and make sure services are all right (Ubuntu)
    • Integrate the old version framework and the latest version OS (Ubuntu)
    • Reinstall the latest version OS and build product environment (Ubuntu)
    • Fix weak points from source code analysis and vulnerability assessment reports
    • Deploy new version product to test and production environment
  • Windows PowerShell (AWS)
    • Write a script to connect to EC2 in a private subnet for customers
    • Write a script to download files from S3 for customers
    • Write a script to upload backup files to S3 and delete local backup files for customers
  • Linux Shell (AWS)
    • Write a script to download files from S3 for customers
    • Write a script to upload backup files to S3 and delete local backup files for customers

Work Experience

Engineer (Information Technology)
International Integrated Systems, Inc.(IISI)

November 2022 ~ June 2023

  • Operation and maintenance for chatbot product
  • Optimize chatbot product
  • Host management (Ubuntu)
  • Database management (PostgreSQL and MongoDB)
  • Openvpn server management
  • Fix weak points from source code analysis and vulnerability assessment reports
  • Assist project manager in satisfying customer's requirement
  • Assist project manager in solving customer's problems
  • Support project manager in technical side
  • Deploy new version product to test and production environment
  • Research monitor and alert
  • Research product CI/CD

Project Engineer (Information Technology)
CHUAN KAI INTERNATIONAL CO.,LTD (CKmates)

June 2021 ~ July 2022

  • Assist the customer in designing, creating, and setting a cloud environment for obtaining PCI DSS Certification
  • Assist the customer in designing, creating, and setting a cloud environment for SAP B1 and HANA Database
  • Assist the customer in designing, creating, and setting a cloud environment for SAP B1 and Microsoft SQL Server Database
  • Assist customers in solving problems
  • Write technical documents for customers and colleagues
  • Be the contact person for projects
  • Organize tables of AWS cloud environment informations for customers and colleagues
  • Research and discuss with colleagues for cloud architecture

Projects (IISI)


Chatbot Product

  • Customization training data
  • Customization functions
  • Management Platform (switch bot status, produce reports,etc)
  • Bot (Q&A, Multi-turn dialogue, element, button)


Optimize Chatbot Product

  • Research the latest version framework
  • Research different component performance
  • Redesign product framework
  • Research automation
  • Research monitor and alert

Projects (CKmates)


Payment Flow Cloud Architecture

  • ECS and EC2 for deploying server
  • Route 53 and ELB for allocating traffic
  • WAF for API and DB security
  • CloudFront for security header
  • CloudWatch for monitoring
  • IAM for necessary permissions
  • VPC for networking



SAP B1 + HANA Cloud Architecture

  • EC2 for deploying HANA DB
  • WorkSpaces for B1 client connection
  • Nginx for web connection
  • S3 for storaging backup and software
  • Lambda for client vpn whitelist
  • CloudWatch for monitoring
  • IAM for necessary permissions
  • VPC for networking 


SAP B1 + SQL Server Cloud Architecture

  • EC2 for deploying SQL Server DB 
  • WorkSpaces for B1 client connection 
  • Nginx for web connection 
  • S3 for storaging backup and software 
  • CloudWatch for monitoring 
  • IAM for necessary permissions 
  • VPC for networking 

Thesis

Building a prediction model for the binding of major histocompatibility complex proteins and peptides 

Projects 01 00@2x

Major histocompatibility complex (MHC) is a set of genes about acquired immune system in vertebrates. The function of MHC is mainly to help vertebrates recognize foreign molecules. There are a lot of issues about MHC in immunity and medical science, such as the susceptibility of autoimmune diseases, adverse drug reaction, and organ transplant. In order to lower the experiments cost, researchers built the prediction model for MHC binding. We built a prediction model for the binding of MHC class I proteins and peptides, named EsMHC, based on machine learning (LibSVM, Weka, and MLP) and a new method of building models: comprehensive strategy. Comprehensive strategy in EsMHC is a strategy to integrate two methods of building models (allele-specific method and pan-allele method) for balancing of strengths and weakness. Moreover, we added MHC Allele encoding developed by ourselves to feature encoding for improving prediction speed and performance. Overall, our model has great prediction performance (AUC = 0.91, ACC = 0.89, F1 = 0.91, and MCC = 0.65). The evaluation metrics of our model are almost higher than other models.

Certifications

AWS Certified Cloud Practitioner

Issued by Amazon Web Services Training and Certification on July 21, 2021

Expires on October 19, 2026

https://www.credly.com/badges/8b96ed58-2ac9-498f-b4ce-667d5d88997d/public_url

AWS Certified Machine Learning – Specialty

Issued by Amazon Web Services Training and Certification on October 03, 2021

Expires on October 03, 2024

https://www.credly.com/badges/9f14f5aa-b46a-4c45-ad3c-fdaf11d8d182/public_url

AWS Certified Solutions Architect – Associate

Issued by Amazon Web Services Training and Certification on November 27, 2021

Expires on November 27, 2024

https://www.credly.com/badges/893a1b78-cbdd-4c33-b254-eeab91d95d3a/public_url

AWS Certified Developer – Associate

Issued by Amazon Web Services Training and Certification on October 19, 2023

Expires on October 19, 2026

https://www.credly.com/badges/1a769f99-b2e6-4a9d-9ba1-019193d0e4bd/public_url

Education

National Chung Hsing University, Master's Degree, Graduate Institute of Biotechnology, 2017 ~ 2020

National Chiayi University, Bachelor's Degree, Department of Plant Medicine, 2013 ~ 2017