Tom Terrier

•Built Airflow data pipelines for operational use powering demand forecasting and business planning with more granular, reliable data • Developed and maintained dbt models on Snowflake • Set up a dedicated Airflow development environment on a GCP server and refactored 15 DAGs to handle environment switching • Built a Streamlit POC on top of Snowflake Intelligence to enable non- technical users to autonomously answer their own data questions

• Dockerized and deployed 6 Machine Learning pipelines performing Quality Checks on medical data and orchestrated on Argo Workflows • Implemented software development best practices on 6 ML pipelines: CI, unit tests and end-to-end tests, alerting, resource optimisation • Developed ingestion scripts to ingest data from SFTP servers to S3 • Built a dbt and Snowflake environment from scratch to monitor and analyze the quality of data received from partner healthcare centers

•Built Airflow ELT pipelines from client data warehouses to Modeo internal data warehouse • Set up the backend infrastructure on GCP (Cloud Run, cloudSQL, Cloud Build, Pub/Sub, Service Accounts, IAM management) via Terraform • Built the backend API (CRUD operations, SSO) • Set up a CI/CD pipeline on Google Cloud Build to automate the API deployment via Google Cloud Run on every push to the main branch

- Development of Python scripts automating metadata collection from YouTube, Instagram, and TikTok APIs into a PostgreSQL database - Development and deployment of an API via fastAPI to execute underlying scripts via Airflow - Dockerization of scripts to run them on AWS Elastic Container Services (ECS) - Set up of Airflow DAGs to periodically collect the metadata through the ECS tasks - Design of the PostgreSQL data model and implementation it via SQLAlchemy and Alembic - Implementation of a Github Workflow to run unit tests and code quality checks on every push to the main branch

- Deployment of Airbyte on Google Kubernetes Engine (GKE) in order to ingest data from MySQL and PostgreSQL into BigQuery - Design of the data model and implementation via dbt - Development of ingestion and transformation pipelines via Prefect - Creation of dashboards on ToucanToco