Socratix

The description states that Socratix is a Socratic study companion built on top of an existing LMS. It integrates into course lessons and allows students to submit free-text answers to questions. The AI layer diagnoses these responses using GPT-5.6 structured outputs, classifying them as CORRECT, SLIP, CONCEPTUAL, or PREREQUISITE_GAP.

Key features include:

• Socratic follow-up questions instead of direct answers
• Refusal of answer-extraction attempts like “just tell me the answer”
• Tutoring confined to instructor-approved syllabus and method
• Generation of next-item practice marked for instructor review
• Logging of next-item correctness to measure learning transfer
• Per-student concept memory storage
• Instructor dashboard showing misconceptions, performance, and generated practice

The system does not run as a free autonomous agent; it uses GPT-5.6 for diagnosis and branching logic is controlled by application code.

Inference The system is built to integrate into an LMS environment rather than function as a standalone tool. This implies it targets educational institutions or platforms already using LMS infrastructure.

The description states that Socratix was inspired by the problem of hidden misconceptions in online courses, where multiple-choice tests fail to show why students are wrong. The goal is to build a tutoring layer that helps students expose reasoning and correct misconceptions Socratically.

Positioning:

• A memory-enabled, agentic Socratic study companion
• Designed for use within an existing LMS
• Focused on introductory inferential statistics
• Aims to unify two normally separate flows: AI-assisted course scaffolding (instructor side) and Socratic tutoring (student side)

Claims:

• The system helps instructors understand what students are actually struggling with.
• It measures learning transfer via next-item correctness.
• It respects instructor control over syllabus and teaching methods.
• It avoids replacing human instruction, instead enhancing it.

Inference The positioning suggests a niche market within higher education or corporate training, where personalized learning and misconception detection are valued. However, no evidence of adoption or usage beyond the demo exists.

The description states that the target course is introductory inferential statistics, focused on sampling distributions, standard error, confidence intervals, and p-value interpretation.

It also implies a dual customer base:

1. Students in introductory statistics courses
2. Instructors who manage those courses and want to track student progress

The system is designed for integration into an existing LMS, suggesting that the primary user group includes educators or course creators using platforms like Canvas, Moodle, or similar.

Inference The ICP likely includes educators teaching foundational quantitative subjects, especially those working within institutions that use LMS platforms. No evidence of specific customer segments or personas beyond this is provided.

Not evidenced.

The description does not mention any pricing model, monetization strategy, or business model. It only describes the technical architecture and functionality of the product.

The system is built using:

• Frontend: Next.js, Tailwind CSS, shadcn/ui
• Backend: TypeScript, Prisma, PostgreSQL, Vercel
• AI Layer: OpenAI Responses API with GPT-5.6 and Zod structured outputs
• Authentication & Payments: BetterAuth, Stripe
• Infrastructure Tools: Arcjet, Neon, pnpm

The AI layer uses GPT-5.6 for diagnosis but does not operate autonomously; the application code controls branching logic.

Key technical decisions:

• Structured outputs via Zod to ensure consistent classification
• Guardrails in application code to enforce Socratic behavior and syllabus adherence
• Integration with existing LMS through schema extensions (Prisma models)
• Use of Codex for rapid development during the hackathon

Inference The architecture suggests a scalable, modular approach that could be extended beyond the demo. However, no evidence of production deployment or scalability testing exists.

Not evidenced.

There is no mention of any revenue, customers, users, or traction metrics beyond the hackathon submission. The project is described as a prototype built in a short timeframe.

Not evidenced.

The description does not reference any competitors or competitive landscape. It does not discuss how Socratix compares to other AI tutoring tools or platforms in the education space.

1. No real-world usage: The product is described as a hackathon prototype with no evidence of adoption or traction.
2. Dependency on GPT-5.6 and structured outputs: Reliance on specific AI models may pose risks if those models change or become unavailable.
3. Guardrail implementation in code: If guardrails are not robust, the system could fail to maintain Socratic behavior.
4. Limited scope: The demo focuses only on introductory inferential statistics; no indication of broader applicability.
5. Unverified claims: All descriptions are self-reported and unverified.

1. What is the current status of the product? Is it being used in any live educational environments?
2. How does Socratix handle edge cases where students provide ambiguous or non-text responses?
3. Has the team tested the system with real instructors and students beyond the hackathon?
4. What are the plans for expanding beyond introductory inferential statistics?
5. Are there any partnerships or integrations with LMS providers already in place?
6. How is the AI model managed and updated? Is there a plan to reduce dependency on GPT-5.6?
7. What would be required to scale this product to support multiple courses or institutions?