Ghaz

The description states that Ghaz is a mobile-first, parent-supervised scavenger-hunt PWA. It uses AI to analyze a room and propose objects for play, with parents reviewing and approving these before they are included in a quest.

• The app supports English and Bahasa Indonesia.
• It includes nine animated pet guides as narrators.
• The experience is designed to be voice-led, with real-time narration via OpenAI APIs.
• Parents control the flow through features like Pause, Parent Help, and End Adventure.
• Visual verification only begins when a parent presses “We Found It”.
• The system uses GPT-5.6 for room analysis and quest generation, validated by Zod and deterministic safety rules.

Note

The author claims to have built the app using React, TypeScript, Next.js, OpenAI APIs, and Supabase. However, no revenue, customer data, or live usage is reported.

The author positions Ghaz as a tool that:

• Transforms familiar rooms into safe, parent-led AI scavenger hunts.
• Enables active family play, not passive screen time.
• Offers parental control over content and safety.
• Is designed to be family-friendly and safe for children.

The project evolved from the idea of helping parents create engaging activities without needing to design them from scratch. It is framed as a solution to the challenge of balancing creativity with child safety in playtime.

Claim vs Fact

The author claims Ghaz makes family play easier, but there is no evidence of user feedback or adoption metrics.

The description states that Ghaz targets:

• Parents who want to engage their children in imaginative play.
• Families with young children, particularly those looking for safe, structured activities.
• Users who value parental supervision and control over digital interactions.

It is not clear if the app is intended for a specific age group beyond general "kids" or whether it targets a niche market like families with toddlers or pre-teens.

Note

No evidence of customer segmentation, personas, or target demographics beyond the general idea of “parents and kids.”

There is no mention of:

• A pricing model.
• Monetization strategy.
• Revenue streams.
• Customer acquisition costs.
• Any commercial intent beyond the hackathon submission.

Inference Since this is a hackathon project, it likely has no business model or pricing structure at this stage.

The author reports:

• Built with React, TypeScript, Next.js, and OpenAI APIs.
• Uses GPT-5.6 Terra for room analysis and GPT-5.6 Luna for visual verification.
• Implements structured output validation (Zod) and deterministic safety rules.
• Supports real-time voice narration via OpenAI Realtime API over WebRTC.
• Uses Supabase for backend services.
• Includes browser-based media handling, with no raw data persisted.

Note

The app is described as a PWA, and the author mentions using Playwright and Vitest for testing. No evidence of scalability, infrastructure, or deployment details beyond development.

The description states:

• This is a hackathon project submitted to the OpenAI 2026 hackathon.
• It was built in one week (Build Week).
• The author used Codex as an implementation partner.
• The app supports fixture mode for testing, but no real-world usage or user data is reported.

Inference No evidence of traction, users, or adoption beyond the author’s own demonstration. No metrics, reviews, or performance data are provided.

The description does not mention:

• Competitors.
• Existing products in the family play or AI scavenger-hunt space.
• Market size or positioning relative to other tools.

Note

The author makes no reference to similar apps or platforms, nor does it describe how Ghaz differentiates from them.

Key risks and red flags include:

• The app is a single-person hackathon project, with no evidence of team, funding, or product-market fit.
• No commercial traction or user feedback.
• No mention of privacy compliance, especially around under-18 data handling.
• The use of GPT-5.6 (not publicly available) raises questions about feasibility and scalability.
• The app is described as a demonstration-only experience, not intended for real child-data use.

Inference Without any evidence of real-world usage, the project may be more of a prototype than a viable product.

1. What is the current status of the app? Is it in production or still in development?
2. Has there been any user testing with parents and children?
3. How does the app handle edge cases, such as ambiguous object detection or user interruptions?
4. Are there plans to expand beyond the current languages (English/Bahasa Indonesia)?
5. What are the long-term goals for Ghaz? Is it intended to be a commercial product or a hobby project?
6. Has the author considered privacy compliance for under-18 users, especially around data retention?