2.56%MAPE
AI Engineer Intern @ CENACE
Sep – Dec 2025 · Coahuila, MX
- Built a Temporal Fusion Transformer that forecasts electrical demand from weather and load data.
- Wrote the end-to-end Python pipeline with Darts: data processing, training and forecasting.
- Reached 2.56% MAPE with interpretable, probabilistic forecasts for power-system planning.
- Python
- Darts
- TFT


