Natural-language intake
Turn a business question into structured entities, assumptions, objectives, and constraints.
Historical filing | July 24, 2025
AI-Powered Supply Chain Intelligence System
A high-level account of U.S. provisional application 63/849,878, which named Kevin Power as inventor and described an interface between language models and structured supply chain decision systems.
This page does not claim a current patent, patent holder, or “patent pending” status.
The only record described here is a provisional application filed July 24, 2025. USPTO guidance says provisional applications are not examined and automatically expire after 12 months unless the applicant takes the required next step. No current public record of a later application is cited on this page.
Filing record
What the historical record establishes
Company filing records identify the application number, title, date, type, and named inventor shown here. A provisional application can establish an early U.S. filing date for subject matter later supported in a qualifying nonprovisional application.
A provisional filing is not examined on the merits and cannot itself issue as a patent. It therefore does not prove novelty, validity, ownership of an issued patent, enforceable claims, or present application status.
Read USPTO provisional guidance- Title
- AI-Powered Supply Chain Intelligence System
- Application number
- 63/849,878
- Filing date
- July 24, 2025
- Application type
- U.S. provisional application
- Named inventor
- Kevin Power
- Current status
- Not asserted on this page
Technical concept
Language as an interface to structured decision tools
At a high level, the filing described a system in which a language-model layer helps users formulate supply chain questions, invokes structured analytical tools, and explains returned scenarios. This summary is descriptive, not a claim chart or statement of legal scope.
Tool orchestration
Route deterministic calculations and optimization work to the appropriate model or service.
Scenario analysis
Compare outcomes when sourcing, routing, capacity, or cost assumptions change.
Result explanation
Return the analytical result with assumptions, trade-offs, and decision context visible to the user.
Product relationship
How the design idea appears in The Trade Lab
The Trade Lab uses a related separation of responsibilities: language models help interpret user intent and product information, while structured services perform tariff lookup, duty calculation, data retrieval, and other bounded operations.
That architectural relationship does not mean the product is patented, that every product feature appeared in the provisional filing, or that any issued claim covers the implementation.
Engineering boundaries
- Authoritative tariff data remains outside the language model.
- Structured tools return inspectable provisions, calculations, and sources.
- Warnings and uncertain facts remain visible for human review.
- The user—not the model—confirms consequential filing decisions.
Why no present-status label appears here
USPTO guidance says a provisional application has a 12-month pendency, is not examined, and does not mature into a patent without additional action. The July 24, 2025 provisional date has passed that 12-month period. Without a current public record establishing a later application or grant, “patent pending,” “patent holder,” and “patented technology” would overstate what this page can verify.
USPTO: types of patent applicationsExplore the working product
See how The Trade Lab combines language interfaces with versioned tariff services and reviewable calculations.
Create free account