• Design a Serverless E2E MLOps Framework Template for classification, forecast and recommender system use cases on rapid AI project productionalization, fully integrated with CICD. Tools involved: GitLab, Docker, Pytest, Google Vertex AI Pipeline, BigQuery, Cloud Function, Pub/Sub, Cloud Storage, Airflow Generator DAG.
• Design a generalized Docker image to pull data from any API to BigQuery. Tools involved: GitLab, Docker, GCR
• Design a generalized Docker image to pull parquet files from GCS to BigQuery. Tools involved: GitLab, Docker, GCR
• Built Airflow ETL Pipelines. Tools involved: GitLab, Airflow, GKE
• Introduce code and docstring coverage to Docker Image structure. Tools involved: Pytest, Interrogate

Program and Processes
• Design and lead a MLOps Standardization Program for entire MLOps team to streamlining the method on productionalizing an AI project.
• Design and lead in building a centralized performance dashboard for all AI projects in terms of business, model and pipeline health. Tools involved: DataStudio, Google BigQuery, Cloud Logging, Pub/Sub, Cloud Function, Bitbucket.
• Revolutionalised team's MLOps ecosystem by building Maxis-specialized and standardized Machine Learning Pipeline, fully integrated with CICD. Tools involved: Bitbucket, Docker, Pytest, Google Vertex AI Pipeline, Secret Manager, BigQuery, Cloud Function, Cloud Scheduler, Pub/Sub, Cloud Storage.
• Built a Git repository as an easy-to-use toolkit for the commonly used DS processes with the implementation of unit testing. Tools involved: Bitbucket, Docker, Pytest.
• Acted as a panelist in Pakar STEM session of Maxis's flagship community program, Maxis eKelas to encourage and support the younger generation who plan on choosing STEM pathway as their own career path.
• Acted as an accessor for Maxis Scholarship Program 2021 to support Maxis Young Talent team in accessing potential scholarship candidates.

Productionalization
• Built a churn prediction pipeline for prepaid users. Tools involved: Refer to standardized pipeline tools.
• Built a churn prediction pipeline for postpaid users. Tools involved: Refer to standardized pipeline tools.
• Built AWS recommender system pipeline for recommending best plans on Hotlink App according to user behaviour and personalisation. Tools involved: AWS SageMaker, AWS Lambda, AWS S3, AWS Glue, AWS EC2, AWS Athena, AWS Quicksight, AWS CloudWatch, AWS EventBridge, AWS SSM.

Innovation
• Drive and involved in building a resume search app that optimize the HR recruitment process. Tools involved: Elasticsearch, Google Cloud Storage, App Engine Flex, Streamlit.

• Automated a customer segmentation pipeline using Azure resources. Tool involved: Azure Synapse, Azure Blob Storage, Azure Databricks, Azure Data Factory, Azure SQL.
• Built a fraud detection pipeline for detecting anomaly transactions & claims for Finance teams. Tools involved: t-SNE, Autoencoder with PyTorch Lightning, Azure Databricks, Azure Blob Storage, Power BI.
• Built a general forecasting pipeline for univariate forecasting. Tools involved: DARTS, Azure Databricks, Azure Blob Storage, Power BI.
• Built a sentiment analysis pipeline for to turn the projects feedback into actionable insights. Tools involved: HuggingFace, Azure Databricks, Azure Blob Storage, Power BI.
• Built a Git repository as an easy-to-use tools for the commonly used DS processes with the implementation of unit testing. Tools involved: GitHub, Docker, Pytest.