Archive position — measured, not model output
1 like on Devpost
506 of the 7,856 archived projects have more likes, and 1,758 share exactly 1 — so this project's #785 place in the like-ranked listing is a tie-break inside that group, not a ranking.
Projects (log scale)
Likes on Devpost. ▲ marks this project's group.
Show the figures
| Likes | Projects | Share of archive |
|---|---|---|
| 0 | 5,592 | 71.2% |
| 1 | 1,758 | 22.4% |
| 2 | 285 | 3.6% |
| 3–4 | 132 | 1.7% |
| 5–9 | 75 | 1.0% |
| 10+ | 14 | 0.2% |
Executive Summary
What the company appears to be
ChatProofs is a self-reported proof engine that evaluates claims using two independent AI models from different vendors. The system is designed to assess whether a claim is supported by evidence, contradicted, or requires further proof — with no model seeing the other’s output.
What changed
The author states they built this tool after observing that 79% of enterprises use both Claude and ChatGPT, interpreting it not as hedging but as a desire for something unnamed. The project was submitted to the OpenAI 2026 hackathon.
Single most important open question — the commercial due-diligence read
Is there any evidence that this system has been used in practice or validated beyond the author’s own testing? The description does not indicate any real-world deployment, revenue, customers, or adoption.
What The Product Actually Is
The description states that ChatProofs is a system that turns a hunch into a proof you can check. It uses two AIs from different vendors:
- One AI (Claude) builds the best honest case for a hunch and finds real sources.
- Another AI (GPT-5.6) sees only the claim and its sources, never the first AI’s output.
The system returns plain-English labels like:
- "A source backs this up"
- "That's a fair leap"
- "Someone needs to prove this"
- "This is actually contradicted"
It enforces that every claim must be backed by a real citation. The system also includes a formalizer (Relay) that structures claims into typed, cited graphs using JSON schema and data contracts.
Inference The product appears to be an experimental AI-assisted fact-checking or validation engine, not a commercial SaaS offering.
Positioning & Claim Evolution
The author claims that the system is built on the idea that humans and agents working together produce more leverage than either alone. They also state that most enterprises use both Claude and ChatGPT, but interpret this not as hedging, but as a desire for something unnamed — implying a gap in current AI tools.
They describe their own product as a way to test claims using cross-vendor verification, which they believe is more reliable than relying on one model alone.
Inference The positioning seems to be that ChatProofs is an experimental tool for validating claims using multiple AI vendors. It is not positioned as a commercial product or service yet.
Target Customer & ICP
Not evidenced.
The description does not state who the intended users are, nor does it describe any specific customer segments or personas. The author’s own use case (validating their own company) is mentioned, but no broader target audience is defined.
Business Model & Pricing Evidence
Not evidenced.
There is no mention of pricing, monetization, or a business model in the description. No revenue streams, subscriptions, or commercial offerings are described.
Technical & Delivery Signals
The system is built using:
- Tools: anthropic, claude, codex, docker, express.js, google-cloud-run, gpt-5.6, javascript, json-schema, mcp, node.js, openai-responses-api, react, typescript, vercel, vite, vitest, web-search, zod
- Architecture: runs on Cloud Run with Secret Manager for keys and rate limits
The description states that:
- The engine enforces a constraint that every citation must be real (URLs must be found by the model).
- A formalizer component (Relay) structures claims into typed graphs.
- Tests are written using Vitest.
- The system logs all outputs, including errors and corrections.
Inference The technical stack suggests a modern, cloud-native development approach. However, there is no evidence of production deployment or scalability beyond the author’s own testing.
Traction & Maturity Signals
Not evidenced.
There is no mention of:
- Customers
- Revenue
- Product usage
- Adoption
- Market traction
The project was submitted to a hackathon and is described as experimental. The author notes that it was tested on their own claims, but there is no evidence of external use or validation.
Competitive Context
Not evidenced.
There is no mention of competitors, market positioning, or how ChatProofs compares to existing tools for fact-checking or AI validation.
Key Risks & Red Flags
- No commercial traction: The system appears to be experimental and not deployed in production.
- Self-reported validation only: The author tested it on their own claims, but no third-party validation is evident.
- Unproven business model: No evidence of monetization or revenue streams.
- Limited team size: Only two members listed, which may limit execution capacity.
- Risk of silent failures: The description notes that the system had bugs where models silently produced identical outputs despite different inputs — a critical flaw in a proof engine.
Diligence Questions To Ask The Founders
- What external use cases or customers have you tested this with?
- Have you validated the system’s accuracy beyond your own testing?
- Is there any plan to commercialize this tool, and if so, what is the business model?
- How do you intend to scale beyond the current experimental setup?
- What are the key assumptions behind the cross-vendor verification approach, and how have they been tested?
Investment/Partnership Verdict
Not evidenced.
There is no evidence of any investment or partnership interest in this project. The description does not indicate any funding rounds, investors, or strategic partners. It is described as a hackathon submission with no commercial traction or market validation.
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.
