<![CDATA[The mobile attendance system at the Secretariat General of the Regional Representative Council of the Republic of Indonesia previously relied on Google ML Kit for face verification services. This approach introduced challenges related to reliance on third-party services, limited configuration flexibility, and institutional control over biometric data. This study aims to develop an internal Face Verification Server using the One-to-One Face Matching method as an alternative to third-party face verification services. The system was developed using the Waterfall software development methodology, covering requirements analysis, system design, implementation, testing, and maintenance. Face verification is performed by comparing a captured facial image with a stored facial embedding associated with an employee identification number. Quantitative evaluation was conducted using 40 verification samples, consisting of 20 genuine and 20 impostor samples, with a threshold value of 0.5. The evaluation achieved 100% accuracy, a False Acceptance Rate (FAR) of 0%, and a False Rejection Rate (FRR) of 0% on the tested samples. The results indicate that the proposed system successfully distinguished matching and non-matching facial images according to the predefined threshold. The internal deployment also reduces reliance on third-party services and enables the face verification service to operate within the institution's internal network infrastructure. This study contributes an independently managed Face Verification Server architecture for mobile attendance systems in a government institutional environment.]]>
