Skip to content

Mira

Mira is GiveCare's AI assistant for family caregivers. She works over SMS, tracks capacity across six domains, screens for benefit programs, reads crisis signals, and follows up when a caregiver asks.

How a turn works

How the runtime works is defined in the SMS runtime contract.

Safety

Boundary decisions are deterministic for exact STOP/HELP responses, pause and restart language, safety holds, obvious emergency signals, and unverified identity-sensitive requests.1 A paused or held caregiver, and an unknown sender, gets the safety reading alone.1

The emergency text says to call 911 now for immediate danger and to call or text 988 for a suicide or self-harm crisis.1 See Crisis Routing.

What Mira offers

Mira learns care context gradually, and asking for a score alone starts no questions.1 She offers an assessment only when code finds the gap and there is no practical request to serve first; the caregiver starts it.1

The one scheduled check-back is the one the caregiver asked for, at the time they named.1

Care as a relationship

Slaughter argues that care is a relationship rather than an activity, and that the relationship is what separates care from a set of automatable services.4 Turkle writes that machines have no friendship to offer.5

Field research and clinical standards

The 2019 research priorities include where technology fits across the caregiving trajectory, adaptive family-centered interventions, real-world complexity and scalability, implementation for diverse populations, and alternatives to RCT and pragmatic trials.2

The UC Davis competencies describe a family-centered approach that recognizes caregivers as integral members of the health team, with shared decision-making and tailored interventions.3

Memory research

MemGPT proposes virtual context management and memory tiers to extend a model's limited context window.6

  1. GiveCare. "GC-SMS Runtime Contract." Exact verified owner projection. Source → ↩↩↩↩↩↩

  2. UC Davis Family Caregiving Institute. "Research Priorities in Caregiving." 2019. Source → ↩

  3. Sexson KE et al. "Interprofessional Family Caregiving Competencies." UC Davis Family Caregiving Institute, 2021. Source → ↩

  4. Slaughter A-M. "Care Is a Relationship." Dædalus 152(1):70, 2023. Source → ↩

  5. Turkle S. "Who Do We Become When We Talk to Machines?" 2024. Source → ↩

  6. Packer C. et al. "MemGPT: Towards LLMs as Operating Systems." arXiv:2310.08560, 2023. Source → ↩