Dr. Diego García

Designed and developed disparate reliable, scalable and cost-effective data pipelines, for batch and event-based streaming applications. Developed on Spark (Scala and Python) and Storm (Java) ingesting from and writing to multiple repositories, such as Kafka, MongoDB, CouchDB and Elasticsearch. Built on Google Cloud Platform and on-premise. Additionally, developed asynchronous APIs in Python (Tornado) to serve results.

Led the development of a recommender system in Python, PostgreSQL and Athena (AWS) for an international e-commerce platform. Developed, among others features, A/B testing and backtesting capabilities to validate alternative treatments. Provided the customer with weekly KPIs to help them understand the results of past actions.

Led the development of data-driven projects based on statistical and predictive models with Hadoop and Spark in Scala, Python and R. In particular, developed a predictive maintenance system for chlorine sensors, defined a customer segmentation based on consumption behavior from large volumes of data and developed a fraud/leak detection system using Automatic Meter Reading data. Ultimately, completed and defended an Industrial PhD thesis.

Developed data cleansing and pre-processing algorithms and predictive models, in Python and C++, for the Automatic and Control Systems research group.

Thesis centered on Data Analytics and Big Data technologies, with the main goal of analyzing the contribution of Big Data technologies to improve decision making in the water sector