CareCatena - DeSci AI caregiver companion for families supporting loved ones with Alzheimer’s

Guild on 0G 2.0
1

Project Name

Care Catena

Project Website

(Frontend running locally / demo deploy expected after hackathon)

Project Brief

What problem you’re solving

Most people with Alzheimer’s are cared for at home by unpaid family members.
These caregivers are overwhelmed, isolated, untrained, and unsupported—yet they perform complex medical, emotional, and behavioral care daily. This takes a major toll on their mental and physical health. At the same time, valuable real-world caregiving knowledge and patterns are lost and never reach clinicians or researchers.

Your solution and key features

Care Catena is an AI companion built to support Alzheimer’s caregivers where they are: at home, in the middle of the work.
It offers:

  • AI Care Companion – context-aware, empathetic guidance for daily challenges (agitation, confusion, communication, routines, safety).

  • Care Memory – a profile of the loved one (diagnosis, routines, triggers, preferences) the AI uses to personalize support.

  • Care Journal – a private, local-first space to capture observations, behaviors, mood patterns, sleep, and day-to-day experiences.

  • Routine & Pill Planner – a lightweight daily helper to reduce decision fatigue.

  • Upload Medical PDFs (Labs, Reports) – planned feature to allow caregivers to store lab results securely and eventually run AI-powered analysis.

  • Future Vision: With family permission, anonymized patterns could help researchers understand early signals, disease progression, caregiver strain, and lifestyle factors at scale and also correlation and patterns of the disease itself.

Target users and market

  • Primary users: Unpaid family caregivers of people with Alzheimer’s or dementia.

  • Secondary users (future): clinicians, researchers, and decentralized science (DeSci) communities analyzing caregiver-reported patterns.
    The global Alzheimer’s caregiver population is over 50 million, with chronic unmet needs and rising urgency.

What makes your project unique

  • Emotion-first AI trained to be calm, supportive, and practical—speaking the caregiver’s language.

  • Contextual Memory: The AI actually remembers the care recipient and the family situation.

  • Local-first privacy: No personal data stored in the cloud; journaling stays on-device.

  • Bridging Care with Science: Real-world caregiving patterns are a blind spot for healthcare—Care Catena aims to unlock them responsibly.

  • Built on 0G: Verifiable, privacy-preserving inference that fits the sensitivity of dementia care.

How will you integrate 0G?

Technical implementation details

  • Backend uses 0G’s Broker SDK to:

    • initialize a wallet

    • acknowledge and deposit to providers

    • request secure LLM inference (LLaMA-based or GPT-OSS-120B)

    • verify TEE proof and signatures

  • Each caregiver chat request is sent as a one-time authenticated query via /services/run with:

    • system prompt

    • conversation history

    • user message

    • care profile memory

The server signs ephemeral headers and handles micropayments automatically.

Which features will use 0G services

  • AI chat companion (current)

  • Personalized recommendations (current)

  • Guided planning and daily routines (current)

  • Future:

    • secure PDF/lab report OCR / extraction

    • biomarker interpretation

    • pattern detection for research

    • secure, on-chain verified inference pipelines

Timeline for integration

  • Now (hackathon MVP): Chat companion + care memory using 0G LLM inference

  • Next 1–3 months:

    • Add PDF upload

    • Add AI-powered lab-marker extraction (OCR + summarization)

    • Add anonymized forum

  • 3–6 months:

    • Enable opt-in data donation to researchers (anonymized)

    • Build longitudinal caregiver dashboards

  • 6–12 months:

    • Launch a decentralized Alzheimer’s caregiving data resource

    • Integrate specialized medical LLMs running on 0G

Expected benefits from using 0G

  • Privacy: Dementia caregiving is emotionally vulnerable—0G avoids centralized data harvesting.

  • Trust: TEE-based verifiability lets families rely on AI output with more confidence.

  • Cost: 0G inference is cheaper than traditional LLM APIs, allowing sustainable scaling.

  • Interoperability: Future DeSci modules and research pipelines can run directly on-chain.

GitHub Repository

Demo / Prototype

Local demo running on: http://localhost:4000 if you deploy from github

Documentation

Social Media

Twitter/X: @yatanbv
Telegram: @yatanbv

1 Like
2

Thanks for the application, it’s cool to see you continuing to work on this after ETHGlobal!

Could you explain in a bit more detail the 4 features that you plan to add the project?

1 Like