Project AutoRepo

The design challenge for ENGIE was to create an NLP-model, which enabled automatic extracting service level objectives and their corresponding value (e.g., ‘minimal room temperature: 21 C°’). This data could then be combined with the actual sensory data which measured the SLO metric. Thereafter, creating a performance report for ENGIE customers which states both the SLA values and the actual measured values. So, in the end the goal was to create an automated software pipeline which was able to:

1. extract necessary information from SLAs,
2. but that together with sensory data,
3. and visualise that in a performance report,
4. which could be directly send to the customer.

The sub questions for this research are:

1. What is NLP, and how does it function?
2. How can NLP be implemented to extract and store KPIs from SLAs?
3. How can sensory data be linked and stored to the extracted KPIs?
4. Can a performance report automatically be generated via extracted KPIs and linked sensory data?