Multiple sclerosis (MS) has heterogeneous trajectories and fluctuating symptoms that are often not captured by intermittent clinic assessments. Digital health tools can provide frequent, low-burden measurements, but clinical translation requires sustained engagement, interpretability, robust governance, and workflow integration. MSPEECH is a single-center, prospective, non-interventional Living Lab study at the MS Center Dresden conducted with ki:elements GmbH to co-create and iteratively refine the Mili smartphone/tablet speech assessment app. The protocol comprises three demonstrator cycles with structured co-design (people with MS and clinicians), implementation within predefined boundaries while keeping core tasks/schedules/security stable, and a 12-week home-based evaluation per version. The battery includes multiple speech tasks (e.g., fluency, reading, "Pa-Ta-Ka," sustained phonation) and a dual-task module combining touchscreen fine-motor tracking with concurrent speech. A dedicated Ethical, Legal and Social Issues (ELSI) board provides continuous oversight. Primary outcomes are usability and acceptability (MAUQ plus qualitative feedback); secondary outcomes include feasibility/engagement, privacy perceptions, report interpretability, clinician workflow fit, and exploratory associations with clinical reference indicators.