NEO:TRACE

The description states that NEO:TRACE is a memory audit tool for long-running AI systems, designed to test two forms of continuity—fact and relationship—using a shared case. The author, a single individual (Neo Samezu), built it using OpenAI Codex and GPT-5.6, with a Streamlit interface. It currently supports a live mode that generates answers and audits them for both fact and relational continuity.

The project is self-reported as a hackathon submission to the OpenAI 2026 hackathon, with no evidence of revenue, customers, or traction beyond its conceptual demonstration. The author claims it tests whether an AI can remember facts but still forget the relationship built through prior interaction. It is unclear if this tool has been used in production or tested on real users.

The description states that NEO:TRACE audits two forms of continuity in long-running AI systems:

• Layer 1 (Fact Continuity): Tests whether the AI uses required current facts, avoids hypotheses, and excludes archived or prohibited claims.
• Layer 2 (Relationship Continuity): Tests whether the AI combines current facts with a response rule learned from a previous failure, user reaction, and repair.

It includes an interface that separates direct fact use, relational evidence, missed requirements, correct exclusions, misuse, and behavior failure. The tool uses GPT-5.6 for baseline generation, NEO:TRACE generation, and structured audit.

The author built it using OpenAI Codex to scaffold a Python and Streamlit application, integrating the GPT-5.6 Responses API workflow and implementing structured output auditing.

Inference The product is described as a conceptual tool for evaluating AI continuity, not a commercial product with a defined market or user base.

The author states that NEO:TRACE began as an attempt to falsify the assumption that “remembering the user is enough.” It positions itself as a test of whether an AI can maintain both factual and relational memory.

It claims to distinguish between:

• Factual correctness
• Relational continuity (i.e., how past interactions shape future behavior)

The tool is described as a way to audit AI systems for their ability to remember facts while also preserving the relationship built through prior interaction.

Inference The positioning suggests a niche, conceptual tool aimed at AI developers or researchers evaluating system behavior, not a commercial product targeting end-users.

The description does not state who the target customer is. It does not name any specific users, personas, or use cases beyond its own demonstration.

Not evidenced No information on who would use this tool, what their role is, or how they might integrate it into their workflow.

The description does not mention a business model or pricing. It is unclear whether the tool is intended to be sold, offered as a service, or used internally.

Not evidenced No indication of monetization strategy, pricing structure, or revenue model.

• The tool was built using OpenAI Codex, GPT-5.6, and Streamlit.
• It uses the GPT-5.6 Responses API workflow.
• It includes structured output auditing.
• It supports both a Live GPT-5.6 mode and a Static Demo.
• The author reports that it was built by a non-engineer using Codex to scaffold, implement, test, and diagnose failures.

Inference The tool is technically feasible but appears to be a prototype or proof-of-concept, not a production-ready product.

The description states that this is a hackathon submission, not a commercial product. It was built by one person (Neo Samezu) and has no evidence of revenue, customers, or adoption beyond its own demonstration.

Not evidenced No data on usage, user feedback, or product maturity beyond the author’s own account.

The description does not mention any competitors or similar tools. It is unclear whether there are existing solutions for auditing AI memory continuity or evaluating relational behavior in long-running systems.

Not evidenced No competitive landscape or market positioning information.

• The tool is described as a single-person hackathon project, with no evidence of team, funding, or traction.
• It is built using GPT-5.6, which may not be publicly available or stable.
• The author is a video creator, not a software engineer, raising questions about long-term maintainability and scalability.
• No evidence of real-world testing, user feedback, or integration into existing AI systems.

Inference The tool is in a very early stage, possibly a prototype with no commercial viability or market traction.

1. What is the intended use case for this tool beyond the hackathon demo?
2. Has it been tested on real users or AI systems?
3. Is there a plan to move beyond a proof-of-concept into a product with a defined market?
4. How does it integrate with existing AI platforms or workflows?
5. What are the long-term technical and business plans for NEO:TRACE?