Gabriel – Python, Django, AWS

Development of an End-to-End Analytics Pipeline Using Satellite Data and Machine Learning.

• Designed and enhanced a scalable, production-ready analytics pipeline leveraging satellite data integrated with advanced machine learning algorithms.
• Improved system scalability and robustness by optimizing task handling and queuing mechanisms in a distributed environment.
• Conducted comprehensive API testing, developed debugging tools, and supported DevOps practices, including CI/CD pipeline implementation.
• Collaborated with a team of quantitative engineers in Munich to drive innovation and deliver high-quality solutions.

A global leader in innovative and sustainable building solutions. Driven by its purpose to build progress for people and the planet, its 60,000 employees are trying to decarbonize buildings while improving living standards for all.

• designed and developed backend infrastructure with Python, Django, REST API, and Celery, together with data pipelines for running AI image analysis for factory anomaly detection;
• designed and developed a frontend interface using Angular for internal stakeholder use.

Energy forecasting platform that deals with production and consumption in key utility areas (electricity, water, gas).

• implemented from scratch the REST API that serves the platform;
• integrated the live ingestion of several weather APIs and the required ETL;
• designed and implemented the ML Ops architecture for running prediction models on AWS using Step Functions, Lambda, Batch Jobs, ECS;
• created the Terraform scripts for provisioning AWS resources;
• managed a team of 3 backend engineers and was responsible for communication between frontend, backend and ML;
• was responsible for the technical due diligence in Seed Round A;
• implemented the terraforming from scratch;
• ensured 90% tests coverage of the backend platform and ML pipelines;
• optimized the resources to run the ML pipelines to minimize time & costs through parallelization;
• delivered the fully functional MVP in 3 months;
• helped the team to raise 2 million in funding in the first seed round.

A platform that aims to reduce the time and costs of genetic disease detection.

• designed and implemented the REST API that serves the platform;
• implemented data engineering ETL pipelines;
• designed and implemented the ML Ops architecture for running prediction models on AWS using Step Functions, Lambda, Batch Jobs, and ECS;
• provisioned AWS resources for the platform;
• was responsible for communication between frontend, backend, and ML;
• was responsible for the technical due diligence in Seed Round A;
• ensured 90% tests coverage of the backend platform and ML pipelines;
• optimized the resources to run the ML pipelines to minimize time & costs through parallelization;
• delivered the fully functional MVP in 3 months;
• helped the team to raise 5 million euros in funding in the first seed round.

It aims to enhance the ability of the marketing team to deploy or update hundreds of thousands of campaigns based on templates.

• reversed engineered legacy code so that the platform managed not only to become code owner but also to extend it with new features;
• designed and built from scratch the second version(MVP) of the platform using GCP services;
• maintained and debugged Angular.js frontend for the platform;
• started developing the new frontend using Angular;
• won the 2019 Drum Search Award for the FoodCat Adwords platform;
• increased the marketing team productivity by at least 100 times.

In order to provide the best experience for consumers, the platform was meant to automatically compensate the client when an unfortunate event happened (delays, canceling, etc.) by sending an email or reimbursing the value to incentivize them to return to the platform. The main features Gabriel worked on there were:

• auto-compensation pipeline;
• dashboard monitor.

• designed and implemented the logic in GCP TaskQueues;
• implemented the monitored dashboard in GCP DataStudio.