ARTIFICIAL INTELLIGENCE INTEGRATION IN TECHNICAL SUPPORT DECISIONS IN THE SOLAR SECTOR Implementation design for GVE Projects Ltd, Nigeria

Abstract Nigeria has the world’s largest electricity-access deficit, and unreliable grid supply drives demand for solar systems and for dependable after-sales technical support. This report examines how artificial intelligence can strengthen the technical support function of GVE Projects Ltd, a Nigerian solar energy company whose public contact model relies on telephone, e-mail and short online forms. It analyses the current support process and compares it with three reference cases: M-KOPA, Huawei FusionSolar smart diagnostics and GVE’s own pay-as-you-go mini-grids. Technical support is treated as a decision system, in which each contact ends in self-resolution, remote fix, field dispatch or specialist escalation. AI is introduced only where it improves one of these decisions and a human can override it. The report proposes a phased, human-supervised system with five architectural layers: WhatsApp-first customer intake, retrieval-augmented knowledge answers, rule-and-model fault triage with a confidence-gated decision rule, and optional telemetry-based predictive maintenance. The design includes an ArchiMate enterprise-architecture model, key performance indicators, a work breakdown structure, a twelve-month plan, a feasibility study, an ISO 31000:2018 risk evaluation and a five-year economic assessment. Because GVE’s internal data were unavailable, the as-is process and the economic model rest on public sources and stated assumptions, and a sensitivity analysis tests them. The report concludes that a narrow, compliance-first deployment under the Nigeria Data Protection Act 2023 is feasible, and that its success depends more on data quality and staff adoption than on the choice of AI model.