OpenAI 2026 hackathon

ChalkTrace - proof-carrying adaptive learning system

Adaptive learning that earns the right to teach: every lesson ships with signed evidence from seven checks, and learners can challenge any claim — disprove it and their path rebuilds.

Solo project by Marc Donovici · 0 likes · 0 comments

Archive position — measured, not model output

0 likes on Devpost

2,264 of the 7,856 archived projects have more likes, and 5,592 share exactly 0 — so this project's #3,198 place in the like-ranked listing is a tie-break inside that group, not a ranking.

Projects (log scale)

1
10
100
1k
10k
05,592
11,758
2285
3–4132
5–975
10+14

Likes on Devpost. ▲ marks this project's group.

Show the figures
LikesProjectsShare of archive
05,59271.2%
11,75822.4%
22853.6%
3–41321.7%
5–9751.0%
10+140.2%
Devpost like counts for all 7,856 archived projects, captured when this archive was built.

Executive Summary

What the company appears to be

ChalkTrace is a self-reported proof-carrying adaptive learning system built as a single-person project by Marc Donovici. It claims to offer an educational platform where learning objects (lessons) must pass seven computed checks before publication and can be challenged by learners, with the system rebuilding the learner's path if a claim is disproven.

What changed

The author states that this is a vertical slice of a larger system, built in response to the asymmetric failure mode of generated tutoring content — where even polished activities may contain errors or misalignments. The product evolved from a focus on "proof-carrying learning objects" and "challengeable adaptation."

Single most important open question

Is there any evidence that this system has been used by learners, educators, or organizations beyond the author’s own development and testing?

Back to contents

What The Product Actually Is

The description states that ChalkTrace is a durable, multi-organization course studio plus a signed sample learning workspace. It allows educators to upload content in various formats (PDFs, IMS Common Cartridge, Moodle, ZIP, QTI/XML, HTML, Markdown, plain text), which are parsed and split into content-addressed source regions.

It introduces the concept of proof-carrying learning objects, where each lesson candidate must pass an eight-check policy before publication. These checks include schema validation, source support, executable correctness, adversarial solvability, hint leakage, accessibility, runtime safety, and educator review.

The system also supports challengeable adaptation: learners can challenge any claim made in a lesson, which triggers a rebuild of the learner's path if the claim is withdrawn.

Evidence

  • The author describes the product as a course studio with upload capabilities for multiple formats.
  • It includes an eight-check policy against candidate bytes.
  • It has a public activity and a separate grading contract.
  • Learners can challenge claims, which causes system-level updates to their learning path.
  • The system uses Ed25519 ProofCards bound to exact hashes.

Inference The system is described as deterministic, with a fail-closed publication gate. It separates proposal from authority and uses cryptographic signatures for trust.

Back to contents

Positioning & Claim Evolution

The author states that the inspiration behind ChalkTrace was to answer a stricter question than “can a model write a lesson?” — what must an AI learning system prove before an activity is allowed to shape a learner's future?

This evolved into two core primitives:

  1. Proof-carrying learning objects
  2. Challengeable adaptation

The author also notes that the product’s own thesis was applied to its own model: GPT-5.6 was used to draft candidates, but they still had to pass all checks and educator review before publication.

Evidence

  • The system is positioned around trust through evidence and interaction.
  • It uses a “proof-carrying” model where each object must prove itself via checks.
  • The challenge mechanism is described as causally changing the future learning path.
  • The author explicitly states that the system applies its own thesis to its own model.

Inference The positioning implies a focus on trust through verifiability, not just content generation. It aims to make trust an interaction, not a disclaimer.

Back to contents

Target Customer & ICP

The description does not clearly define a target customer or ideal customer profile (ICP). However, the author notes that the first adoption target is an engineering or science course team running high-enrollment assessment practice.

It also mentions that the concrete two-week rollout is written up in docs/pilot-playbook.md.

Evidence

  • The first adoption target is described as an engineering or science course team.
  • A pilot playbook exists for a specific use case (high-enrollment assessment practice).

Inference The system seems aimed at educators and institutions who want to build or use high-quality, verifiable learning content — particularly in STEM fields.

Back to contents

Business Model & Pricing Evidence

There is no evidence of pricing, business model, or revenue streams in the description. The author does not mention any monetization strategy or customer acquisition plan.

Evidence

  • No mention of pricing.
  • No indication of a commercial model (e.g., SaaS, freemium, licensing).
  • No information on how users pay or what they get for it.

Inference The business model is not evidenced, and the project appears to be in an early development stage without traction or monetization plans.

Back to contents

Technical & Delivery Signals

The system is built using:

  • Codex, ed25519, fastapi, mcp, python, react, typescript, web-crypto
  • It includes a 2,048-case executable verifier tied to both Python and TypeScript implementations.
  • The system uses SQLite for durable storage, with every account, course, draft, publication, attempt, challenge, and audit event stored.
  • It implements cryptographic ProofCards and DecisionCards.
  • A release gate recomputes deterministically.

Evidence

  • The system is built with a mix of Python, TypeScript, React, FastAPI, and web-crypto.
  • It includes an 8-check policy with exact-hash evidence.
  • It uses Ed25519 for signing and cryptographic verification.
  • A deterministic release gate exists that recomputes results.

Inference The system is technically sophisticated, with strong emphasis on cryptographic integrity, deterministic verification, and evidence preservation.

Back to contents

Traction & Maturity Signals

There is no evidence of traction or adoption beyond the author’s own development. The project is described as a single-person effort, and no customers, users, or revenue data are provided.

Evidence

  • Team size is listed as 1.
  • No mention of users, customers, or real-world deployment.
  • No data on usage, engagement, or performance metrics.
  • No evidence of product-market fit or user feedback.

Inference The system is in a pre-traction phase, likely at the prototype or MVP stage. It has not yet been validated with external users or organizations.

Back to contents

Competitive Context

There is no mention of competitors or market context in the description. The author does not reference existing platforms or tools in the adaptive learning or educational technology space.

Evidence

  • No competitor names, products, or market positioning are mentioned.
  • No discussion of how ChalkTrace compares to other systems.

Inference The competitive landscape is not evidenced, and it’s unclear whether this project addresses a gap or overlaps with existing solutions.

Back to contents

Key Risks & Red Flags

  1. Single-person development: The system is built by one person, which raises concerns about scalability, maintenance, and long-term viability.
  2. No traction or user feedback: There is no evidence of real-world usage or adoption.
  3. Unproven business model: No indication of how the product will generate revenue or attract users.
  4. High technical complexity without validation: The system is technically complex but lacks evidence of being tested or used in practice.
  5. No external validation or third-party audits: The project is self-reported and unverified.

Evidence

  • Team size: 1.
  • No user data, revenue, or adoption metrics.
  • No mention of external testing or feedback.

Inference The project is high-risk due to lack of traction, unclear business model, and limited validation.

Back to contents

Diligence Questions To Ask The Founders

  1. What are the actual use cases for which this system has been tested?
  2. How does the system handle edge cases in content ingestion or challenge resolution?
  3. Has there been any real-world feedback from educators or learners?
  4. What is the plan to scale beyond a single developer?
  5. Are there any plans to monetize or commercialize the product?
  6. How does the system ensure that the source material used for training models is valid and peer-reviewed?

Back to contents

Investment/Partnership Verdict

Not evidenced

The project description provides no evidence of revenue, customers, traction, or a clear path to monetization. It is a self-reported prototype with strong technical underpinnings but no commercial validation.

Confidence Low This is a pre-traction, pre-commercialization product, built by one person and not yet tested in the market. Any investment or partnership would be highly speculative at this stage.

Back to contents

Source

Submitted to the OpenAI 2026 hackathon on Devpost. Project home on DevPost.

The analysis above was generated by a language model from the project's own one-line description. It is not independent research and contains no verified traction, revenue or customer data.