Rung

The description states that Rung is a differentiated-instruction platform. It allows teachers to set up digital classrooms where students take a short diagnostic test. Based on results, it builds individualized practice plans for each skill that needs improvement.

Key features include:

• AI tutor offering three levels of help: nudge, hint, and guided step
• Live heatmap showing where each student stands
• One-click actions from the evidence
• AI-drafted mini-lessons for small groups with shared learning gaps
• YouTube video guidance included in lessons

The platform uses Next.js, Supabase, and GPT 5.6 (as per author's declaration), with a clear separation between what GPT helps with and what the app decides itself.

Inference: The product is described as an educational tool aimed at helping teachers tailor instruction to individual student needs, using AI for both tutoring and lesson planning.

The authors position Rung as a way to replicate the teaching style of Ms. Collins — a teacher who identifies exactly where each student is stuck, meets them there, and guides them without giving answers directly.

They state:

• “Rung finds the exact sub-skill each student is missing”
• “Tutors them there”
• “Turns the class's gaps into tomorrow's lesson plan”

This positioning implies a shift from traditional one-size-fits-all instruction to personalized learning at scale, using AI as an enabler.

Inference: The platform positions itself as a solution for educators seeking to improve student outcomes by addressing individual learning gaps more effectively than standard curricula allow.

The description states that Rung is intended for teachers and students in high school or similar educational environments, particularly those who want to implement differentiated instruction.

It also mentions:

• Teachers setting up digital classrooms
• Students taking diagnostics and receiving personalized practice plans
• Mini-lessons tailored to small groups with shared learning gaps

Inference: The primary customer is likely K–12 teachers, possibly in under-resourced or remote educational contexts where personalized attention is scarce.

Not evidenced. The description does not contain any information about pricing models, monetization strategies, subscription plans, or revenue streams.

The team built Rung using:

• Next.js
• Supabase
• GPT 5.6

They also mention:

• Clear architecture and contracts.md files
• Use of Codex for building data flow and UI
• Validation of model outputs before showing them
• Prevention of answer leaks in hints
• Safe fallbacks when AI features fail
• Test suite of 230 tests across 36 files

Inference: The team shows technical discipline, particularly around AI integration and safety measures. However, no evidence indicates deployment, scalability, or production readiness.

Not evidenced. There is no mention of:

• Customers
• Revenue
• Usage metrics
• Product adoption
• Any form of user feedback or pilot programs

The project was submitted to a hackathon and has no documented traction beyond its own description.

Not evidenced. The description does not reference competitors, existing solutions in the edtech space, or market positioning relative to other platforms.

• Unproven pedagogical effectiveness: No evidence that AI tutoring improves learning outcomes.
• AI safety concerns: While they claim to prevent answer leaks and use fallbacks, there is no independent validation of these safeguards.
• Prototype status: Built for a hackathon; no indication of product-market fit or long-term viability.
• No commercial data: No revenue, customers, or usage data provided — all claims are self-reported.

1. What specific learning outcomes have been observed in pilot classrooms using Rung?
2. How does the platform validate that its diagnostic assessments accurately identify student skill levels?
3. Can you provide examples of how the AI hints and guided steps differ from direct answer provision?
4. Have you tested the system with real teachers and students, or is it still in experimental phase?
5. What are your plans for scaling beyond a hackathon prototype?