Optimization Assurance · Local browser prototype
A second look at your facility plan.
Check demand, assignments and capacity. Reproduce the cost. Keep a review you can inspect and share.
Study files are processed on this device. They are not uploaded or stored by this site.
Select your study
Choose all six files together. Stable IDs connect clients, candidate sites and proposed assignments.
study.json · clients.csv · sites.csv · costs.csv · selected.csv · assignments.csv
- No study selected.
Or explore a generated example
Examples are synthetic. They do not represent a customer study or actual savings.
Supported model and file format
Weighted sum of assignment costs; exactly-k or at-most-k open sites; required sites; and one whole-client assignment to an open site. Blank capacity means unlimited. Required sites need not be filled.
Up to 512 clients and 128 candidate sites. Eight MiB per file, sixteen MiB combined. Decimal and fraction strings stay exact. Additional constraints receive an unsupported-model result.
Costs and demand units must agree. Physical measurements, travel routes, fixed opening costs, split demand, fairness and time windows need separate models.
Full input contract and examples ↗Review the evidence
Select a study or load an example to begin.
Every accepted cost has a checked plan.
Your review will show capacity and assignment failures, exact costs, and the limits of any quality claim.
Findings
No optimization solver runs in this browser. The review checks your submitted plan and a nearest-site alternative; capacities can make that alternative infeasible.
Input evidence and provenance
Supplied provenance is not independently authenticated. Review sign-off: not supplied.
Keep this review
Download a standalone report or the evidence bundle containing the exact original files, checks and engine identity. Exports include your study data.
Copy exports if your browser blocks downloads
Copy the evidence into a new file ending in .json, or copy the standalone report into a new .html file.
What a checked result means
A feasible plan meets the constraints you declared. The lower bound relaxes site-count, required-site and capacity restrictions. Its difference from a checked cost bounds suboptimality; it does not reveal the actual optimum unless both values match.
These checks do not certify geographic data, the mathematical source theorem or a business outcome. This prototype has finite validation and explicit limits. The local Python tool can separately request and check conventional solver proposals.
Inspect the calculations and validation →