The energy fault detection system for smart buildings is integrated with cloud-based Databricks Workplace, this Advanced High-Performance Computing Environment for Big Data applications running on Apache Spark. Avoiding a Smart Building Diagnostics as a Service system and establishing a centralized design allowed for the rapid creation and implementation of this cloud-based system within one calendar year. Thus, I also managed to develop a 24/7 uninterrupted data provision workflow using all the Databricks integrational features that included data fetching, cleaning, and analysis of energy consumption, including forecasting and anomaly detection. Currently, the indicated system is capable of monitoring and detecting faults for 96 administrative buildings located on an active military base. The different activities of the system are executed within an average of 14 min. It also secures the first interface to the cloud data lake through SFTP and BLOB file system secure protocol drivers for automated first-phase data loading. These processes were integrated into collaborative notebooks within the pipeline using PySpark—a reliable Python binding for Apache Spark. The pipeline has been well controlled and set up in terms of workflow and continues to meet the deployment tasks. Regarding the architecture of our system, this paper outlines it and distinguishes it from previous attempts at diagnosing smart buildings; it also describes the technology stack of the data pipeline of the big data analytics solution, as well as key steps for BLG’s management and updating in the Indian context.