Deciphering cell signaling pathways is essential for advancing our understanding of basic biology, disease mechanisms, and the development of innovative therapeutic interventions. Recent advancements in multi-omics technologies enable us to capture cell signaling information in a more meaningful context. However, omics data is inherently complex-high-dimensional, heterogeneous, and extensive-making it challenging for human interpretation. Currently, computational tools capable of inferring cell signaling pathways from multi-omics data are very limited, underscoring the urgent need to develop such methods. To address this challenge, we developed Incytr, a method that facilitates the efficient discovery of cell signaling pathways by integrating diverse data modalities, including transcriptomics, proteomics, phosphoproteomics, and kinomics. We demonstrate Incytr's application in elucidating cell signaling within the contexts of COVID-19, Alzheimer's disease, and cancer. Incytr successfully rediscovered known subpathways in these diseases and generated novel hypotheses for cell-type-specific signaling pathways supported by multiple data modalities. We illustrate how overlaying Incytr-identified pathways with prior knowledge from biomarker and small molecule drug databases can be used to facilitate target and drug discovery. Overall, as we demonstrated here, with the use of simple natural language processing AI models, these pathways could serve as a discovery tool to deepen our understanding of cell-cell communication semantics and co-evolution.