Success stories • Oil & Gas
New data platform with analytics and machine learning
The implementation of Microsoft Fabric enables this company in charge of coordinating the generation, transmission and distribution of power in Argentina to capitalize on the value of its data and even generate new monetization alternatives.
About the client
Argentine company in charge of the operation, supervision and administration of the Wholesale Electricity Market (MEM) in the country. Founded in 1992, it coordinates the generation, transmission and distribution of electricity, ensuring an efficient and reliable power supply. It is also responsible for planning and scheduling the daily operation of the electricity system and settling economic transactions in the wholesale market.
With a strong focus on transparency and efficiency, it works in collaboration with generators, distributors and large users to optimize the operation of the national electricity system. It relies on advanced technologies and innovative management practices to maintain the stability and quality of the power supply. It also plays a key role in implementing energy policies and promoting the use of renewable sources in Argentina, contributing to the sustainable development of the energy sector.
A modern data platform in the cloud
Client’s need
To have a new modern data platform in the cloud that allows to adopt or develop traditional analytics solutions and advanced machine learning models.
– Develop a traditional analytics solution in the cloud (with layers for Raw, Bronze, Silver and Gold).
– Ability to monetize on the data available to the company (data sharing).
The client faced several challenges in terms of the evolution of the technological platform, which needed to be solved in order to achieve a better use of the data.
Solution
• A new cloud data platform was designed and developed.
• The solution was based on Microsoft Fabric.
• Ingests were designed for a SQL Server database where energy meter information is evidenced and another Oracle database with meter measurement data (KPI).
Se diseñaron las ingestas para una base de datos de SQL Server donde se evidencia información de medidores de energía y otra de Oracle con datos de mediciones (KPI) de los medidores.
• A lakehouse was created where the information from both databases is stored using a delta layer population system such as the Raw, Transform and Mars layers.
• In the lakehouse service, this information is processed and a first processing is performed in which a history of the data is made available in the Transform layer.
• Then the data is processed with the generation or updating of a data model generating 38 tables in which a layer is created in the Mart layer.
• In the datawarehouse a process of incremental loads was designed for the transactional tables coming from the Oracle database to the DataFlows Gen 2.
