A Blueprint That Needs a Builder
We've done the hard design work. The architecture is defined, the safety logic is specified, and 20+ behavioral test cases are written. What we don't yet have is working code — and that's what we're looking for help with.
Founding Lead Architect Search: We are seeking a technical partner to harden the Three-Model Orchestration Layer and move from specification to a working clinical pilot.
What we have
- Full spec for the Safety Moderator
- Pydantic schemas for all LLM outputs
- Initial "Safety Bench" test cases (JSON)
- Clinical validation protocol (CARE-SAT)
What we need
- LangChain/LangGraph implementation
- Python/FastAPI microservice hardening
- Cloud deployment (AWS/GCP) architecture
- Security-first data logging design
The Ethics of AI in Vulnerable Populations
We believe AI must be a caregiver-support tool, not a replacement for human connection. Our governance model ensures that technology remains bounded by clinical ethics.
The CARE-SAT Safety Standard
We maintain a versioned public standard (v1.0 currently adopted) that governs all system behaviors:
- Non-deception: Explicit refusal to impersonate family members or make false relational claims.
- Caregiver-in-the-Loop: 100% of PLWD-facing features are caregiver-controlled and session-bounded.
- Clinical Boundaries: Strict refusal of medical, legal, or financial advice with automated human-handoff triggers.
⚕️
Geriatrics & Neuro
⚖️
Bioethics & AI Safety
🤝
Caregiver Advocacy
The Search: We are currently recruiting Clinical and Ethics Advisors to join our board and help steward these standards during our clinical expansion.
A Clinical Moat in a Commodity Market
AI companions for dementia are proliferating. None have addressed the safety and clinical accountability barrier required for real adoption. Lucid Bridge is building that barrier first.
$355B
Annual US Cost of Dementia Care
$30K
Seed Milestone Ask
8 Weeks
Pilot Duration
What's Been Built
- AI Companion Safety Standard v1.0 — published and versioned
- CARE-SAT Pilot Protocol — complete 8-week feasibility study design
- Three-model technical architecture — specs and pseudocode
- Hybrid organizational structure — for-profit IP + 501(c)(3) nonprofit research arm
This is a seed ask to generate the first safety evidence in this category — data that unlocks larger grants ($200K–$500K) and eventual B2B partnerships.
The Founding Vision
Lucid Bridge is currently a physician-led solo initiative moving from proof-of-concept to clinical validation. We are seeking "Founding Partners" who want to build the future of safe AI.
Brian Harris, MD
Founder & Clinical Lead
A physician focused on sleep and cognitive health (EusomniaMD), bridging the gap between clinical ethics and generative AI. Dr. Harris provides the "Ground Truth" for the CARE-SAT safety standards.
"Our success is not measured by 'time spent on the app,' but by the quality of the human interaction that occurs after the session ends. We are building the AI that knows its place."
Who We Are Looking For
We are seeking high-leverage founding partners to move from prototype to clinical pilot:
- Lead Systems Architect: To harden our 3-model orchestration layer.
- Ethics & HCI Advisor: To ensure senior-accessible, non-deceptive design.
- Research Coordinator: To oversee the Redwood City clinical pilot.