In Central Brooklyn, Black and Latino caregivers in safety-net pediatric clinics are often asked to complete intake forms that do not reflect how they naturally communicate. Traditional tools require precise reading and numeric responses, leading to incomplete data, added staff burden, and less time for behavior-change counseling. This proof-of-concept study introduces a literacy-aligned, voice-ready conversational intake workflow designed to match how caregivers actually speak, including rough estimates, narrative responses, and everyday language. To prepare for real-world use, we conducted a four-step simulation cycle (Design, Simulate, Evaluate, Refine), testing the workflow across 31 scripted scenarios before involving any families or clinical systems. Results: All 31 scenarios were completed successfully. The workflow correctly interpreted and routed responses 95 percent of the time, and two design gaps were identified and resolved. No patient data were used. This work represents a pre-implementation phase supporting future evaluation of caregiver usability, workflow integration, and staff adoption in safety-net clinical settings.