Javier – SQL, AWS, Python

It's a task tracker used for personal needs.

• Developed back-end functionality using a microservice architecture.
• Built front-end components with FastHTML.
• Implemented notifications using the ntfy.sh service.
• Wrote unit and integration tests to ensure system reliability.
• Integrated LLMs to generate day-at-a-glance summaries.

It's a wrapper around ntfy.sh service to foster and ease the implementation of shell notifications for any process. This package is designed explicitly for self-hosting services.

• Developed the core code wrapper.
• Designed system interactions and workflows.
• Open-sourced the project code.

It's an intrapreneur venture aimed at digitalizing a car dealership.

• Developed back-end architecture using a microservices approach.
• Built front-end interfaces to support back-office operations.
• Integrated with instant messaging providers.
• Integrated with credit score services, including Equifax.
• Deployed backend services through CI/CD pipelines, maintaining 99.999% uptime.
• Wrote unit, integration, and end-to-end tests to ensure system reliability.

It's an open-source Python version of the Spicy IDs system.

• Developed an implementation of the Spicy ID framework in Python.
• Open-sourced the project code.
• Wrote comprehensive tests to ensure functionality and stability.
• Deployed the package to PyPI.

The project aimed to improve the quality and speed of data in the company’s data lake and data warehouses for a fintech platform serving businesses.

• Deployed the Dagster orchestrator for data workflow management.
• Developed DAGs to automate and manage data processing tasks.
• Integrated Dagster with DBT for data transformation and modeling.
• Implemented process triggers to streamline pipeline execution.
• Integrated the system with all relevant data sources and sinks.

It's a data lake development for a fintech platform serving businesses.

• Built a medallion-tiered data lake on Google Cloud Storage (GCS).
• Deployed a CDC solution using Datastream on GCP, integrating MySQL and PostgreSQL sources across multiple clouds.
• Implemented data lake pipelines with DBT across all layers.
• Built enriched One Big Tables to serve internal customer needs.

It's a credit scoring pipeline for a fintech platform serving businesses.

• Developed a scoring algorithm and refactored code using clean architecture.
• Built a pub/sub process for near-real-time and batch entity processing.
• Wrote unit and integration tests with Pytest.
• Deployed the solution on AWS.

It's a data lake development and maintenance for a utilities company from Chile.

• Created and deployed SSIS pipelines and artifacts to source data from SQL Server.
• Built and deployed PySpark pipelines to integrate data from additional sources.
• Developed a data lake and integrated it into a SQL Server database.
• Designed and populated multiple data warehouses.
• Maintained and optimized the data lake architecture.

It's a data lake for a utilities company to enhance its operations and facilitate cross-selling across multiple channels.

• Developed a data lake integrating data from multiple databases and platforms.
• Established a governed data management process based on the DAMA framework.
• Deployed batch and near–real-time data acquisition pipelines.
• Utilized BigQuery scheduled queries to create and maintain tables and views in a structured manner.
• Deployed PySpark pipelines to perform large-scale data transformations.

It's an ML containerization and deployment for a government agency to analyze its workload and enable operational efficiencies.

• Containerized an API-based ML model using Docker.
• Orchestrated Docker containers with Docker Compose.
• Deployed the solution on a remote platform using Make and Bash scripts.
• Developed and fine-tuned ML models using MLflow and Python.

The project focused on building a daily predictive model to forecast customer energy consumption for a utilities company using time series analysis.

• Developed PySpark data pipelines to feed input to ML algorithms.
• Built and deployed predictive models using Facebook’s Prophet framework.
• Optimized Spark code to reduce runtime and improve performance.
• Developed and deployed multiple ML models using MLflow.