SPI Knowledge Miner

Company: SPI Knowledge Miner

Self-reported basis: The description is entirely self-reported by the author, unverified, and contains no evidence of revenue, customers, traction or operational history.

What it appears to be: A decision-intelligence workspace that uses synthetic data and deterministic logic to return either an “Evidence Passport” (for supported questions) or a clear refusal (for unsupported ones). It is built as a React + Vite web app with no external integrations or production workflows.

What changed: The project was submitted to the OpenAI 2026 hackathon, indicating it is a prototype or proof-of-concept.

Most important open question: Is there a viable business model or path to monetization beyond the isolated demo?

The description states that SPI Knowledge Miner is a governed decision-intelligence workspace. It claims to:

• Turn a business question into an auditable evidence contract
• For supported questions, return an Evidence Passport, which includes:
  • Audience
  • Period
  • Method
  • Permitted claims
  • Limitations
  • A deterministic local receipt
• For unsupported or disallowed requests (e.g., predictions or individual-targeting), it returns a clear no-match and refuses to generate an invented answer

The product is described as a self-contained React + Vite site, running entirely on synthetic, local aggregate fixtures. It does not integrate with any external data sources, production systems, or real-world workflows.

Inference: The product appears to be a prototype built for demonstration purposes, likely in a hackathon setting, and not yet deployed in a commercial environment.

The author states that the product is designed to avoid ambiguity in business decision-making by ensuring that answers are grounded in evidence. It positions itself as a tool that:

• Prevents teams from relying on AI-generated summaries without supporting data
• Offers a clear, auditable path for decisions
• Refuses to answer unsupported questions

The tagline — “A governed decision-intelligence workspace that finds the evidence an answer can stand on—and refuses the questions it cannot support” — reinforces this positioning.

Inference: The product is positioned as a governance-first AI assistant, focused on preventing overconfidence in AI outputs and ensuring transparency in decision-making. It is not a general-purpose AI tool but a specialized one for evidence-based business decisions.

The description does not specify the target customer or ICP (Ideal Customer Profile). It only states that the product is intended to help business teams who have data but lack clarity on what evidence supports their decisions.

Inference: The likely target audience includes business decision-makers, product managers, data analysts, or compliance officers in organizations that prioritize governance and auditability of AI use. However, no explicit customer segment is defined.

There is no evidence in the description of a business model or pricing structure. The product is described as a public demo with no login, billing, or external integrations. It uses synthetic data and does not appear to be connected to any real-world systems or workflows.

Inference: No commercial model is evident. The project appears to be a non-commercial prototype, likely built for a hackathon.

The product is built with:

• Technology stack: React + Vite
• Deployment: Self-contained, synthetic, local-only
• Logic: Deterministic matching and refusal behavior
• Testing: 75 automated checks across behavior, privacy, claim limits, and build quality
• Security/Privacy: No person-level data, no external integrations, no production workflows

The description also mentions:

• Codex built the judge-safe public experience
• GPT-5.6 implemented Evidence Passport mechanics and test suite
• The majority session ID is 019f7ced-9f46-7673-b26d-d0ab13401503

Inference: The technical architecture is minimalist, built for demonstration, not production. It uses synthetic data and deterministic logic to simulate behavior, but lacks real-world integration or scalability.

There is no evidence of traction, customers, revenue, or adoption beyond the public demo. The project is described as a hackathon submission, with no indication of:

• User base
• Product usage metrics
• Customer feedback
• Iteration history
• Production deployment

The product is explicitly stated to be isolated and synthetic.

Inference: No traction or maturity signals are evident. It is a proof-of-concept, not a developed product.

There is no evidence of competitors or market context in the description. The author does not reference any existing tools, platforms, or markets for decision-intelligence or AI governance.

Inference: No competitive landscape is described. It is unclear whether this product addresses an existing market gap or introduces a novel concept.

• No commercial viability: The project is a demo with no monetization or production use case.
• No real-world data or integrations: The system only works with synthetic data and does not connect to live systems.
• Unproven business model: No evidence of how the product would generate revenue or scale.
• Limited scope: It is built for a specific use case (business questions) and lacks general-purpose AI capabilities.
• No customer or user feedback: The project has no real-world testing or adoption.

1. What is the intended commercial model, and how does it scale?
2. How would this product integrate with existing enterprise systems or data sources?
3. What are the key assumptions about governance that need to be validated in a real-world setting?
4. Is there any plan to move beyond synthetic data into production use cases?
5. How do you envision the product evolving from a hackathon prototype to a commercial offering?