//assets.cakeresume.com/assets/contentarea/contentarea-2d64eca8dae2a1ffb31e0fb1df887bbad072c7c19734d7d30fa14884bbcc299a.css

* Implement and study the MLOps in Cinnamon.
* As Data Engineer, have over 3yr experience in python
* Data processing with Hadoop & Spark under E-commerce, Fintech, Music streaming domain
* From scratch, Deploying Scalable data pipeline in microservice with Kubernetes and Airflow
* Knowing how to develop & deploy in GCP & AWS
* Had the experience to implement recommendation system, marketing analysis dashboard

National Central University

Machine learning engineer

Jayce Li

Cinnamon

Software Engineer

* Develop the general worker chassis and apply it to 90% Cinnamon project.
* Own the integration project to promise AI system delivery to the Client successfully.
* Implement MLOps to speed up and centralize models
* Implement SaaS model service to help retrain the model, register models, and update prediction service.
* Package AI pipeline as a docker image to bridge the SE and AIR work.
* Implement an auto-labeling system to release human labor resources.

Leading a team with 10 members in Martech:
* As a tech lead to collect and integrate the team’s option then rapidly implement a prototype to discuss with the client to get feedback.
* As a scrum master, make sure everything should be fine in the scrum
* Building serverless services to replace Cloud composer in 2 weeks.
* Using Cloud composer (Airflow) to build over 30-thousand ETL and retrieve data from Google Ads API
* Using Bigquery as data storage to speed up data developing
* Learn & use Laravel to build a frontend with the team website in a week.
* Use Swagger to generate readable API
* GCP Data Engineer

CloudMile

Data Engineer


Fintech with 10 team member:
* Applying agile development to speed up software development.
* Building ETL to Kubernetes with Apache Airflow and Spark in GCP.
* Design airflow-dynamic-etl framework on GitHub
* Design & implement backup & recovery situation
* Developing SparkAccess lib to reduce the possibility of data corruption
* Building CI/CD with Jenkins, SonarQube, Docker. 

E-commerce A with 4 team member:
* With scikit-learn, try to find the score of the item-user pair.
* With Solr to develop the marketing analysis dashboard

E-commerce B with 2 team member:
* Using tf-idf to find important feature based on the product description

Music-Streaming with 10 team member:
* Using Spark to find the top K user as the representative base on the user's behavior.

E-commerce C with 6 team member:
* Using Spark to label the user tag base on the user's buying log.
* Using HBase, Spark, Node.js to Implement a master-slave, cacheable, scalable API server 


Share