We present the assessment of a diffusion-dominated mean field axisymmetric dynamo model in reproducing historical solar activity and forecast for solar cycle 25. Previous studies point to the Sun's polar magnetic field as an important proxy for solar activity prediction. Extended research using this proxy has been impeded by reduced observational data record only available from 1976. However, there is a recognised need for a solar dynamo model with ample verification over various activity scenarios to improve theoretical standards. The present study aims to explore the use of helioseismology data and reconstructed solar polar magnetic field, to foster the development of robust solar activity forecasts. The research is based on observationally inferred differential rotation morphology, as well as observed and reconstructed polar field using artificial neural network methods via the hemispheric sunspot areas record. Results show consistent reproduction of historical solar activity trends with enhanced results by introducing a precursor rise time coefficient. A weak solar cycle 25, with slow rise time and maximum activity $-14.4 \%$ ($\pm 19.5\%$) with respect to the current cycle 24 is predicted.