<p>Plasma profile reconstruction is essential for analyzing experimental results from the KSTAR tokamak, as accurate measurements of electron temperature (<i>T</i>_e), ion temperature (<i>T</i>_i), and electron density (<i>N</i>_e) are critical for understanding plasma behavior and validating theoretical models. Existing reconstruction tools at KFE, such as the OMFITprofile module and an EPED model-based Markov chain Monte Carlo (MCMC) method, provide reliable results, but require substantial manual input and computational time, making them unsuitable for high-throughput experimental campaigns where plasma shots are spaced only 10&#xa0;min apart. To address these limitations, we developed a machine learning (ML)-based framework for rapid plasma profile reconstruction. The method builds upon the EPED model parameterization and predicts six key profile-shaping parameters directly from diagnostic measurements, enabling accurate reconstruction of pedestal-shaped profiles characteristic of H-mode plasmas. Compared to the traditional MCMC-based fitting, the ML approach achieves a significant reduction in computation time while maintaining high fidelity to experimental data. This development enables fast, automated, and reliable plasma profile reconstruction, supporting near real-time analysis and improving the efficiency of experimental campaigns. Preliminary results demonstrate that the proposed method achieves comparable accuracy to conventional techniques while reducing reconstruction time from several minutes to seconds.</p>