Bridge

The description states that Bridge is a real-time conversation aid for individuals with cerebral palsy. It uses:

• OpenAI Realtime API (gpt-realtime-2.1) for live audio transcription and speech
• GPT-5.6 to generate response cards based on the question
• A React frontend with WebRTC, an Express server for API key handling, and Codex for development

Key features include:

• Real-time transcription of clinician questions
• AI-generated response cards (yes/no, pain scale, feelings, locations, open answers)
• Patient selection via eye gaze, switch, or keyboard
• Instant speech output of selected card
• Session memory that builds on prior turns
• Optional loading of patient context for personalization

The system is designed so that:

• AI generates the choices but never makes the choice
• Cards are generated in parallel to avoid blocking live session
• Voice model remains silent unless triggered by a card selection
• Card generation includes validation and revision loops to ensure quality and safety

Evidence Self-reported, unverified. No revenue, customers or adoption data.

The author states that Bridge was inspired by the mismatch between understanding and interface in communication for people with cerebral palsy. The core claim is:

• “AI proposes; they choose.”
• The system aims to improve accessibility by removing bottlenecks in communication.
• It is built with ethical constraints, such as no inferred intent, no automatic replies, and opt-in context.

The positioning appears to be a tool for clinicians working with individuals who have limited speech capabilities. It is framed as an assistive technology that uses AI responsibly.

Evidence Self-reported claims about intent and design philosophy. No external validation or market positioning data.

The description states that Bridge targets:

• People with cerebral palsy who understand spoken language but cannot respond verbally
• Clinicians (e.g., speech-language pathologists) interacting with these patients
• Users requiring assistive communication tools (AAC)

It also mentions:

• Eye-gaze and switch hardware support
• Single-switch users
• Integration with clinical data systems (FHIR, audit logs, encryption)

Evidence Self-reported. No evidence of customer base or user segmentation beyond the stated target group.

There is no mention in the description of:

• Revenue model
• Pricing structure
• Monetization strategy
• Customer acquisition plans

The project is described as a hackathon prototype with no indication of commercial viability or business development.

Evidence Not evidenced.

Key technical elements include:

• Use of OpenAI Realtime API (gpt-realtime-2.1) and GPT-5.6
• WebRTC for audio handling
• React frontend, Express backend
• Codex for development workflow
• Semantic voice activity detection to transcribe-only mode
• Structured output with bounded plan → validate → revise agent loop
• Session memory that treats past selections as historical statements

The system is designed to be:

• Low-latency
• Safe (no automatic replies)
• Accessible (keyboard, switch, eye-gaze navigation)
• Agentic but bounded

Evidence Self-reported. No evidence of scalability, performance metrics, or production deployment.

There is no evidence of:

• Revenue or ARR
• Customers or users
• Product adoption or usage data
• Market traction or growth
• Prior funding or investor interest

The project is described as a prototype built in a hackathon. It has not been commercialized or deployed beyond the demo stage.

Evidence Not evidenced.

There is no mention of:

• Competitors
• Market landscape
• Existing solutions in AAC or assistive communication tools
• Differentiation from other platforms

The description does not provide any competitive analysis or positioning relative to existing tools.

Evidence Not evidenced.

Key risks and red flags include:

• Unverified claims: All descriptions are self-reported, with no independent verification.
• Prototype-only status: No evidence of real-world deployment or clinical validation.
• No commercialization path: No indication of how the product will be monetized or scaled.
• Limited team size: Only one founder listed (PradeepKumar Vasudevan).
• Ethical constraints as design challenge: While intentional, this may limit functionality or scalability.
• Dependency on proprietary APIs: Reliance on OpenAI’s Realtime API and GPT models introduces risk of changes or access limitations.

Evidence Inferred from self-reported description. No external data to validate or contradict.

1. Has Bridge been tested with actual users (people with cerebral palsy) in clinical settings?
2. What is the current status of clinical validation and user feedback?
3. Are there any plans for clinical data integration beyond demo JSON files?
4. How do you plan to scale beyond a single developer prototype?
5. Have you considered alternative AI models or APIs if OpenAI changes its offerings?
6. Is there any interest from healthcare providers or AAC organizations in piloting this?
7. What is the path to commercialization and monetization?