Cadence

Cadence is a self-reported real-time communication companion for people with ALS and other AAC users, built as a prototype for the OpenAI 2026 hackathon. The product aims to support participation in fast conversations by offering prepared replies, conversation initiation, and offline resilience, all while maintaining user control over speech output.

The project is described as an early prototype, not a medical device or emergency service. It uses browser-based live transcription, GPT-5.6 for reply generation, and local storage for privacy. The team consists of two members and the product is built with Next.js, TypeScript, Tailwind CSS, OpenAI APIs, and MediaPipe.

Key commercial due-diligence read: There is no evidence of revenue, customers, or traction beyond the prototype phase. The self-reported description lacks any data on user adoption, usage metrics, or commercial viability. The product’s positioning as a communication tool for people with ALS may imply a niche market, but there is no indication of how this would scale or monetize.

Single most important open question: Is there a viable path to market beyond the prototype phase, and what are the barriers to adoption by end users or partners?

The description states that Cadence is a real-time communication companion for people with ALS and other AAC users. It supports:

• Live captions using browser Speech Recognition API
• Prepared replies generated via GPT-5.6 with structured JSON output
• User control over speech, including editing, rejecting, or saving suggestions
• Initiation of conversation topics (not only reaction)
• Offline resilience through local storage and device voice
• Voice options including OpenAI TTS or instant device speech
• Accessible input methods: touch, keyboard, scanning, eye-gaze focus

The system is built with Next.js, TypeScript, Tailwind CSS, OpenAI APIs, MediaPipe, localStorage, and Vercel.

Inference: The product appears to be a web-based application designed for accessibility, with an emphasis on real-time interaction and user agency.

The description states that Cadence focuses on participation in the whole conversational turn, not just faster typing. It emphasizes:

• Timing, choice, and connection
• Be ready before the moment passes
• Choose language that sounds like the person
• Repair a wrong suggestion or misunderstanding
• Hold the floor when more time is needed
• Initiate something meaningful, not only react

It positions itself as different from other AAC tools by focusing on real-time communication and user control, rather than just text composition speed.

Inference: The positioning evolved from addressing the limitations of existing AAC tools to enabling full conversational agency for users with ALS.

The description states that Cadence is intended for:

• People with ALS
• Other AAC users
• Users who communicate slowly and need support in fast conversations

It also mentions that it supports caregivers, speech-language pathologists, and assistive-technology professionals through optional partner tips.

Inference: The primary ICP appears to be individuals living with ALS or similar motor impairments, with secondary users being caregivers and professionals involved in AAC support.

There is no evidence of a business model or pricing strategy in the description. The project is described as an early prototype and not a commercial product.

Inference: No revenue streams or monetization plans are evident from the self-reported description.

The system is built with:

• Next.js 14 App Router
• TypeScript
• Tailwind CSS
• OpenAI APIs (GPT-5.6, TTS)
• Browser Web Speech API
• MediaPipe Face Landmarker
• localStorage for local memory
• Upstash Redis for rate limiting
• Vercel for deployment

It includes features such as:

• Mock mode for offline behavior
• Structured outputs and API route separation
• Consent checks, input caps, CSP, rate limiting, and server-side secret handling

Inference: The technical stack suggests a modern, web-based, privacy-conscious application with modular design and offline capabilities.

The description states that Cadence is an early prototype, not a medical or emergency tool. It was submitted to the OpenAI 2026 hackathon and has no evidence of revenue, customers, or adoption beyond its own self-reporting.

Inference: There are no traction or maturity signals beyond the prototype stage. No data on user engagement, retention, or product usage is provided.

The description mentions that many communication tools help compose text faster but Cadence focuses on participation in the whole conversational turn, not just typing speed.

It references research and studies related to AAC, ALS, accessibility, and communication-participation, but does not name specific competitors or market players.

Inference: The competitive landscape is implied to include existing AAC tools, but no direct comparison or differentiation from known products is made.

• No commercial traction or revenue: The product is described as a prototype with no evidence of adoption or monetization.
• Unverified claims: All features and benefits are self-reported without independent validation.
• Prototype status: Not a production-ready tool, nor a medical device.
• Limited scalability: No indication of how the solution would scale beyond a single prototype.
• Privacy and safety: While privacy-first design is emphasized, there is no evidence of compliance with regulatory standards or clinical trials.

Inference: The lack of real-world usage data, commercial viability, and product maturity raises significant concerns about its readiness for market entry.

1. What are the key user feedback loops from co-design sessions with people who use AAC?
2. How do you plan to validate that the AI-generated replies are relevant and accurate in real-world settings?
3. Is there any clinical or research validation of the product’s effectiveness in improving communication outcomes?
4. What is the path to market beyond the prototype, including regulatory considerations and user adoption strategies?
5. Are there any partnerships or pilot programs with caregivers, speech-language pathologists, or assistive-technology organizations?
6. How do you intend to monetize this tool if at all?