Compass

This analysis is based entirely on the self-reported, unverified description provided by the project author. No external verification, revenue data, customer names, or third-party sources are available. All claims are labeled as "the description states" and should not be interpreted as facts without corroboration.

The description states that Compass is a bilingual learning companion for students working through geometry exercises in GeoGebra. It supports both voice and text interaction, offers progressive hints, and can guide students through structured missions such as the Varignon theorem investigation.

It uses:

• GeoGebra (under a non-commercial license)
• OpenAI GPT models, including gpt-5.6-terra, gpt-5.6-luna, and gpt-realtime-2.1
• Next.js, TypeScript, React, Playwright, Vercel

The system allows students to:

• Work directly in GeoGebra.
• Receive AI guidance without doing the thinking for them.
• Have their constructions verified by deterministic checks within the application.
• Interact with a conversational AI that interrupts only under specific conditions.

It is described as a one-person project built during a hackathon, not yet part of any commercial product or platform.

Not evidenced: No information on actual users, usage metrics, or deployment beyond the author’s own testing and demo.

The description states that Compass aims to:

• Help students work through exercises without taking the thinking away.
• Combine GeoGebra as a space for experimentation with an AI companion that understands the environment.
• Offer progressive help, not just solutions.
• Provide deterministic verification of mathematical correctness, distinct from AI-generated feedback.

It positions itself as:

• A guided practice tool, not a replacement for teachers.
• An educational AI companion, not a generic chatbot.
• A specialist investigation platform, currently focused on geometry.

The author notes that Compass does not claim to:

• Correct every exercise.
• Understand every subject deterministically.
• Replace a teacher.

Inferred: The positioning appears to be evolving from a personal idea into a prototype for educational AI use, with clear intent to limit AI authority and maintain student ownership of work. However, no evidence of prior versions or market positioning beyond this single submission.

The description states that Compass targets:

• Students working through geometry exercises.
• Teachers who can prepare activities using a draft editor.
• Educators interested in AI-assisted learning tools.

It is designed for:

• School-level students (e.g., Varignon theorem investigation).
• Use cases involving geometry, with potential for expansion into other subjects.

The author mentions:

• A public one-click demonstration available without login or classroom code.
• A teacher space where educators can prepare and edit activities.
• A limited classroom pilot using pseudonyms and rotating codes.

Not evidenced: No data on actual student demographics, teacher profiles, or adoption rates. The target customer is inferred from the described use case rather than confirmed by external evidence.

The description states:

• Compass is built under a non-commercial license for GeoGebra.
• It does not produce grades or store personal data beyond what’s necessary for the learning experience.
• No pricing model, monetization strategy, or revenue streams are mentioned.
• The author intends to make source code available to the GeoGebra community.
• If it wins a hackathon prize, 20% will be used to fund API credits for a school.

Not evidenced: No business model, pricing plans, or commercial arrangements are described. The project is presented as a prototype with no indication of monetization or scalability beyond its current form.

The description states:

• Compass uses Codex extensively for engineering and review.
• Built using Next.js, TypeScript, React, and Playwright.
• Uses OpenAI GPT models (gpt-5.6-terra, gpt-5.6-luna, gpt-realtime-2.1).
• Implements a closed set of semantic actions to prevent unrestricted access to GeoGebra.
• Includes an agent harness that controls AI intervention timing and prevents over-helping.
• Supports real-time interruption handling, canceling responses when constructions change.
• Uses WebRTC for low-latency voice interaction.

Inferred: The technical architecture suggests a strong focus on controlled, deterministic interaction between AI and student activity. However, no evidence of scalability, infrastructure, or production deployment is provided.

The description states:

• Compass was built during an OpenAI 2026 hackathon.
• It includes a public one-click demonstration.
• A teacher space exists for preparing activities.
• A limited classroom pilot has been conducted with pseudonyms and rotating codes.
• The author plans to test in real classrooms with teachers and students.

Not evidenced: No data on user engagement, retention, or adoption beyond the author’s own testing. No evidence of revenue, customers, or product-market fit is available.

The description does not mention competitors or similar products directly.

It implies Compass is a specialist tool focused on geometry and real-time AI tutoring in educational environments.

Not evidenced: No competitive landscape, market positioning, or comparison to existing tools in the edtech space is provided.

• Single-person development: The project is built by one individual, which raises questions about scalability and long-term maintenance.
• Limited scope: Currently focused only on geometry investigations; no indication of broader subject support.
• No commercialization strategy: No pricing model, monetization plan, or business model described.
• Unverified claims: All features and functionality are self-reported without independent validation.
• Dependency on external tools: Relies heavily on OpenAI models and GeoGebra, both of which may change or become unavailable.
• Lack of traction data: No evidence of user adoption, feedback loops, or performance metrics.

Inferred: The lack of any commercial or user-facing signals suggests a high risk of failure to scale beyond the prototype stage.

1. What is the current status of the GeoGebra license and its implications for future development?
2. How does the author plan to transition from a hackathon prototype to a scalable product or service?
3. Are there any plans to expand beyond geometry into other subjects?
4. Has the author tested Compass in real classrooms? If so, what were the outcomes?
5. What are the technical and financial constraints of scaling this tool for wider educational use?
6. How does the author intend to monetize or commercialize Compass?
7. Are there any partnerships or institutional support beyond the hackathon?