All challenges
MVPEnergyNDA required
Detecting consumption anomalies
Tassili Énergie· Verified organisation
Automatically identify faulty meters or fraud on our network.
The problem
Anomalies are spotted manually, often months after the fact.
Expected impact
Significant non-technical losses and customer disputes.
Expected results
Daily anomaly score per meter with explanation.
Success criteria
Accuracy above 80% on a sample validated by our field teams.
Trades needed
Artificial intelligence
- Machine learning
Data
- Data science & statistics
Skills sought
- Machine Learning
- Séries temporelles