The scarcity of effective biomarkers and therapeutic strategies for predicting disease onset and progression in neurodegenerative diseases such as Alzheimer's disease (AD) is a major challenge. Conventional drug discovery approaches have been unsuccessful in providing efficient interventions due to their 'one-size-fits-all' nature. As an alternative, personalised drug development holds promise to pre-select responders and identifying suitable indicators of drug efficacy. In this exploratory study, we have established a pipeline with the potential to guide patient stratification studies before clinical trials. This pipeline uses 2D and 3D in vitro models of monocyte-derived microglia-like cells (MDMi) from AD and mild cognitive impairment (MCI) patients, and matched healthy control (HC) individuals. By profiling cytokine responses in these models using multidimensional analyses, we have observed that the 3D model offers a more defined separation of profiles between individuals based on disease status. While this pilot study focuses on AD and MCI, future investigations incorporating other neurodegenerative disorders will be necessary to validate the pipeline's findings and demonstrate its broader applicability.