EnerQ

EnerQ is a multi-agent AI energy system for commercial facilities. It was piloted on the Sultan Qaboos Complex for Youth, Culture and Entertainment in Salalah, Oman. The system uses five specialist agents that are coordinated, not one monolithic script. They observe consumption, investigate waste, simulate candidate interventions against a facility Digital Twin, decide, act, and report. EnerQ acts and explains itself in fluent Arabic or English.

The five agents are ObserverAgent, DiagnosticAgent, SimulationAgent, ActionAgent, and ReportsAgent. Each agent owns 2-3 of the pipeline's nine stages. The pipeline stages are Observe, Detect, Investigate, Generate Solutions, Simulate, Compare, Decide, Recommend, and Verify. The agents are independently replaceable.

The decision engine uses deterministic multi-criteria scoring: 60% savings, 30% risk, 10% comfort. The reasoning layer uses an LLM behind an OpenAI-compatible API, with a deterministic fallback if the LLM is unreachable. The Digital Twin is an industrial-monitoring-style dashboard over a facility physics model (HVAC, lighting, plug loads, solar, 3 real zones). The system is bilingual throughout, with every page, agent explanation, and knowledge-base chunk in both Arabic and English.