Background and objectives
Major energy player present in 70 countries.
As part of the transformation plan for the Finance Department of energy marketing entities, the program aims to automate, make more reliable and facilitate the closing process, financial forecasts and pricing.
The program covers only “energy margin” activities (top of P&L) and therefore excludes OPEX.
Our mission
Project PMO: roadmap definition, comitology coordination, technical/functional facilitator and industrialization of the project approach.
Product owner: definition of product vision, expression of business needs, backlog and product development supervision.
Data management: initialization of data governance associated with the finance data domain.
Results
Set up an AWS base automatically fed by data providers
Top P&L modeling in DataIku to feed budget, rolling forecst and best-estimate processes
Data Quality Monitoring dashboard implementation (DataIku + Power BI)
AI's contribution to historical data recovery projects
Automated data extraction:
Pattern recognition : automatic identification and extraction of relevant data from various sources, even if unstructured.
Data cleansing and transformation:
Anomaly detection: identification and correction of errors or inconsistencies in data, thus improving the quality of transferred data.
Data normalization: automatic conversion of data into a standardized format suitable for the new system.
Data migration:
Intelligent planning: optimize the migration plan by identifying the best strategies and sequences for transferring data.
Transfer automation: reduce manual intervention by using AI algorithms to manage data transfer.
Validation and verification:
Quality control: checking that the data transferred is complete and accurate.
Automated testing: running tests to ensure that the data works correctly in the new system.
Predictive analysis:
Problem forecasting: anticipate potential problems during migration and propose proactive solutions.
Results
Time savings: Significantly reduce the time needed to extract, clean and transfer data.
Greater accuracy: fewer human errors thanks to automation and continuous verification.
Efficiency: Smoother processes and less disruption to day-to-day operations.
Adaptability: Ability to manage large and diversified data volumes.
