REHEARSE

The description states that REHEARSE is an interactive, evidence-grounded simulation engine, not a quiz generator or generic chatbot. It compiles knowledge into a causal scenario graph and allows learners to make decisions with consequences.

It includes:

• A learner workflow: work order, symptom confirmation, inspection, hypothesis formation, verification.
• An instructor analytics layer: decision path, observations, evidence used, diagnosed misconception, confidence levels.
• A deterministic engine: no AI controls scenario state or scoring; model output is a candidate only.

The author describes one deep automotive case as the initial demonstration. The system uses synthetic data (vehicle specs, prices, times) to simulate real-world diagnostic scenarios without using proprietary manufacturer data.

It is built with:

• Next.js, React, TypeScript, Zod
• OpenAI Responses API with Structured Outputs
• Playwright for browser checks
• Cloudflare Workers

Not evidenced: No mention of revenue, customers, pricing, or actual deployment beyond the author’s own testing and demo video pipeline.

The author states:

"Textbooks tell. Videos show. REHEARSE lets you practice."

This positions REHEARSE as a practice-based learning tool, distinct from traditional instructional formats. It emphasizes evidence-grounded reasoning and diagnostic decision-making in vocational contexts.

Claims include:

• The system diagnoses misconceptions behind reasoning
• Learners can recover from wrong decisions with consequences
• Instructors get visibility into how learners thought

The author also states that the product is a reusable education engine, not limited to automotive diagnostics, although the first case is automotive.

Inference: The positioning suggests REHEARSE targets vocational instructors and onboarding teams looking for safe, repeatable practice environments. However, no evidence of market traction or adoption exists.

The author identifies:

• Primary audience: Vocational instructors and workshop onboarding teams
• Use case: Practicing rare, costly, or unsafe-to-recreate decisions before real-world application

They also state:

"A pilot should measure whether learners choose more discriminating tests, reduce unnecessary simulated cost and time, recover from misconceptions, and need less instructor intervention on repeated cases."

This implies a target customer persona focused on training outcomes and efficiency gains.

Not evidenced: No mention of actual customers, partnerships, or feedback loops. The description does not name specific institutions or users.

The author states:

"The public judge experience is intentionally keyless and deterministic so it remains fast and reliable."

This suggests a free-to-use public demo with no explicit pricing model described.

There is no mention of:

• Subscription tiers
• Licensing models
• Enterprise or institutional use cases
• Revenue streams

Not evidenced: No business model, pricing structure, or monetization strategy is provided.

The system is built using:

• Next.js, React, TypeScript, Zod
• OpenAI Responses API with Structured Outputs
• Playwright for browser checks
• Cloudflare Workers

Key technical features include:

• 45 deterministic unit and integration tests
• 12 desktop/mobile development browser checks
• 6 optimized-production browser checks
• Input/output validation, origin protection, rate limits, kill switch
• Prompt-injection defenses
• Dependency, license, accessibility, secret-leak audits

The author notes that:

"Codex accelerated the build" and helped move from concept to working vertical slice.

They also state:

"Approved scenarios are cached, and model calls are reserved for generation and adaptive wording rather than ordinary learner transitions."

Inference: The system is built with a strong emphasis on determinism, security, and validation, suggesting an engineering approach that prioritizes reliability over generative AI control.

The author describes:

• A single, deep automotive case
• A demo video pipeline
• A private repository with immutable submission tag and integrity manifest

They also note:

"I ran a sanitized owner proof against the exact gpt-5.6-sol model."

However, there is no mention of:

• Customer adoption
• User feedback or usage metrics
• Product iteration history
• Real-world deployment

Not evidenced: No evidence of traction, user base, or product maturity beyond the author’s own development and testing.

The description does not reference:

• Direct competitors
• Similar tools in the education or simulation space
• Market positioning relative to other vocational training platforms

Not evidenced: No competitive analysis or market context provided.

• No revenue, customer or adoption data: The product is described as a prototype or demo with no evidence of real-world use.
• Unproven learning outcomes: While the author describes intended pilot metrics, no actual results are shared.
• Single-person team: The project is built by one individual, raising questions about scalability and long-term maintenance.
• Limited scope: The system is demonstrated only in automotive diagnostics; its broader applicability is untested.
• AI dependency with strict boundaries: While the author says AI is used only for compilation and adaptive wording, this may limit innovation or adaptability.

Inference: Without evidence of traction or feedback, REHEARSE remains a concept-stage product. The lack of external validation raises risk that it may not meet real-world needs.

1. What is the current status of the pilot testing? Are there any early results?
2. How do you plan to scale beyond one domain (automotive)?
3. What are your plans for monetization or commercialization?
4. Have you received feedback from vocational instructors or training teams?
5. How do you intend to validate that learners actually improve their decision-making through use?
6. What is the long-term vision for the product beyond this hackathon submission?