Hepatocellular carcinoma (HCC) typically develops on a cirrhotic background, where conventional serum biomarkers such as alpha-fetoprotein (AFP) provide inconsistent performance. We implemented an integrated serum intact glycoproteomics strategy that profiles N-linked and O-linked glycopeptides in parallel. This strategy couples prevalence-aware filtering with multi-algorithm feature selection to prioritize glycoform-resolved markers distinguishing HCC from cirrhosis. Serum from HCC (n = 20) and cirrhosis (n = 20) underwent high-abundance protein depletion, proteolysis, glycopeptide enrichment, and high-resolution LC–MS/MS. Intact glycopeptides were identified with Byonic/Proteome Discoverer and filtered to a high-confidence quantitative matrix requiring ≥70% feature presence.