A major issue with the scientific research information management system is the wrong appraisal of results, which has prompted a reform of scientific research management in response to the widespread use of Internet technology. The conventional management mode cannot accomplish the efficient management level in the scientific research information management system, and the assessment is inappropriate. Consequently, this paper suggests an AI-powered network algorithm for assessing and evaluating technological and scientific advancements in management. Firstly, the deep learning theory is utilized to assess scientific researchers, and the indicators are split according to the management evaluation needs to eliminate the interference elements in management evaluation. Following this, a management evaluation scheme is developed and the outcomes of this scheme are thoroughly examined using the principles of deep learning theory, which has been applied to the assessment of services related to technical and scientific innovation. When compared to more conventional management models, MATLAB simulations reveal that intelligent network algorithms outperform them in terms of management evaluation efficiency and accuracy when it comes to services related to technological innovation and scientific research.