House: Continuity With Jurisdiction

The company appears to be a solo-developer project named "House: Continuity With Jurisdiction", submitted as a hackathon entry to the OpenAI 2026 hackathon. The author describes it as a private AI workspace focused on user control over conversational continuity, with features like editable memory, explicit search (Rummage), and privacy boundaries. It is built using React, TypeScript, Supabase, Cloudflare Workers, and OpenAI tools.

What changed: This is a self-reported project description from a single developer, submitted as part of a hackathon. There is no evidence of prior development, funding, or commercial traction beyond the author’s own account.

The single most important open question: Is there any evidence that this product has moved beyond the prototype stage, or that it has been tested with users outside of the developer's own environment?

Note: This analysis is based entirely on the self-reported project description provided by the caller. No external verification, funding history, customer data, or revenue information is available.

The description states that House is a private AI workspace focused on continuity and competence in conversations. It includes:

• Spaces with distinct conversational context
• Editable, resident-controlled Continuity memory
• Never Remember boundaries
• Explicit provider, model, and reasoning selection
• Conversation import and review
• Attachment lifecycle and deletion controls
• Rummage — a search tool that finds exact words across retained conversations and returns canonical source results

It is described as a React and TypeScript application, using Supabase for authentication and data persistence, and Cloudflare Workers for AI request handling.

Inference: The product appears to be a prototype or proof-of-concept built during a hackathon. It is not evidenced to have been deployed in production or used by others beyond the developer.

The author positions House as an AI workspace where continuity stays under user control, contrasting it with AI assistants that remember too little, wrong things, or use personal context without clarity.

Key claims:

• Continuity is useful but should not become invisible or uncontrollable.
• Residents decide what belongs in memory, what must never be remembered, and which model should answer.
• Rummage allows exact-word search across conversations and returns canonical sources.
• The system avoids silent provider substitution or fallbacks.
• It emphasizes privacy boundaries, provenance, and inspectability.

Claim vs. Fact: These are self-reported claims about intent, positioning, and design philosophy — not evidence of adoption, traction, or performance.

The description states that House is built for residents who want to control conversational continuity in AI interactions. It is framed as a private workspace, implying individual users rather than enterprises or teams.

Inference: The target is likely a single user or small group of users who value privacy and control over their AI-assisted conversations. No evidence of segmentation, personas, or customer types beyond the developer’s own use case.

There is no evidence in the description of any business model or pricing structure. The author does not mention monetization, subscriptions, licensing, or any commercial offering.

Not evidenced: No information on how the product would be sold, who pays, or what revenue model is intended.

The project is built with:

• Frontend: React, TypeScript
• Backend: Supabase (authentication and data), Cloudflare Workers (AI layer)
• Tools: GitHub for version control, Codex for development and adversarial review, Vitest for testing

Key technical claims:

• Rummage supports authenticated cross-conversation retrieval, result ranking, source navigation, privacy boundaries, provider isolation, lifecycle testing, and protections against leaking sensitive data.
• The developer used Codex extensively for implementation, testing, and adversarial review.
• The system avoids unsafe abstractions by choosing to not merge a prototype that required database-level transaction authority.

Inference: The technical stack is modern and well-suited for a prototype. The use of Codex suggests an experimental or AI-augmented development process. However, no evidence of production deployment or scalability.

The project is described as a hackathon submission, built during Build Week. It is not evidenced to have any users, customers, or real-world usage beyond the developer’s own testing.

Not evidenced: No data on user adoption, retention, revenue, or product-market fit. The author explicitly states that no traction data exists beyond their own account.

The description does not mention any competitors or direct market comparisons. It is framed as a solution to problems with AI assistants' memory and context handling, but no specific competing products are named or described.

Not evidenced: No competitive analysis or positioning relative to existing tools in the AI workspace or conversational AI space.

• Solo developer project: The team size is listed as 1. This raises questions about long-term maintenance, scalability, and product development capacity.
• Prototype-only: The project is described as a hackathon submission with no evidence of production use or user testing.
• No commercialization plan: No mention of monetization, pricing, or go-to-market strategy.
• Overly technical focus: The emphasis on database-level transaction authority and concurrency control suggests a high level of engineering complexity that may not align with early-stage product needs.

Inference: The project is likely in an early prototype phase, with no clear path to commercial viability or user adoption.

1. What was the actual scope and timeline of the hackathon project?
2. Have you tested this with any users outside of your own environment?
3. Are there plans to move beyond a prototype into a product that could be used by others?
4. How do you intend to monetize or scale this product?
5. What are the key assumptions about user needs and behavior that underpin the design choices?