Noeun Kim

Serving API

• Developed model serving APIs using Django, FastAPI, and Apache Solr resulting in concurrent request issue resolution.
• Converted Python-based serving API to Rust, resulting in improved performance and efficiency.

Data Engineering

• Analyzed order log data through exploratory data analysis (EDA) to prepare them for model development and training.
• Developed data pipelines using Spark and Airflow, Google Kubernetes Engine, facilitating efficient data processing on distributed cluster nodes.
• Implemented CDC for Real-time Data Ingestion via Apache Flink, enabling direct data writing to DocumentDB.
• Managed data lineage and quality with DBT for maintaining data integrity throughout the pipeline.
• Collaborated with data scientists, and product owners to enhance communication.

Model Management

• Utilized MLFlow for efficient management and selection of model versions.
• Leveraged Grafana, Python, and sqllint to utilize feature and model score dashboard for monitoring performance.

• Developed DNA analysis platform serving APIs using Python, PyTorch, and Sun Grid Engine
• Refactored MSA API to optimize performance
• Achieved migration of DNA databases from on-premise infrastructure to AWS cloud.
• Managed CI/CD pipeline using AWS CLI, Jenkins, and GitActions.

• Developed skills in conducting on-device deep learning model compression research
• Created a platform for monitoring GPU performance

• Developed and maintained data analytics protocols and documentation standards
• Analyzed complex genome sequence data to identify anomalies, trends, and risks, providing valuable insights for service