Commercial priorities and the first product experiment
The catalog contains 41 software/business dossiers, all 372 family assessments and 719 inventoried manuscripts. The leading hypothesis remains one shared evidence and model-assurance platform. There are 42 runnable utilities, 21 documentary contracts and 32,820 passed checks in 67 saved reports. These are research and integration evidence; no buyer demand, profitable deployment or independent proof acceptance has been established.
Build first: an Optimization Assurance Workbench for facility-planning reviews by logistics/operations consultancies. This is the selected first experiment within the shared platform, not a validated market winner. A concrete generated case shows a solver FEASIBLE plan independently certified optimal by an exact dual witness. Source runtime remains unestablished, so use conventional proposal engines and useful independent checking. First product specification, revised dossier.
Build and earning queue
| Priority | Workflow and candidate buyer | Concrete capability now | Evidence needed before a commercial claim |
|---|---|---|---|
| 1 | Facility-planning assurance for logistics/operations consultancies | Stable-ID CSV intake, separate strict and weighted/capacitated plan checks, exact bounds, optional existing-solver proposals and HTML/JSON review exports | Permissioned real study, useful recurring review gaps, willingness to pay and delivery cost |
| 2 | Aggregate-data law/constraint review for statistical engineering teams | Exact tiny laws, fixed-event comparison, explicit bound conditioning; live SciPy agreement | Valid actual experiment design, material errors caught and advantage over existing practice |
| 3 | Allocation certificates and flexibility review for operations teams | Existing-solver matching witness checks, attaining cardinality bounds and tiny counts | Real mandatory constraints, independently accepted certificates and actionable delivery value |
| 4 | Bounded-flow scenario/model review for planning teams | Binary signed-bound reduction, tiny exact arc-vector diagnostics and checked existing-solver proposals | Useful scenario law/units, practical large-support backend and planning decisions improved |
| 5 | Scientific/quantum numerical assurance for domain teams | Exact geometry, correlation, spectral, aliasing and polynomial-bound controls | Domain fidelity, numerical/program bridges and buyer budget |
| 6 | Sparse permutation framework research for data-access teams | Shared-switch replay, decimal-wire provider, tiny-law and entropy checks | Universal numerical threshold, compact storage, realistic sparse/batched workload and fair-bit/PRG distinction |
| 7 | Three-machine scheduling assurance for narrow unit-job workflows | Direct witness checks, tiny optima/deadline evidence and live CP-SAT adapter | Repeated value beyond existing solver output and accurate real-model correspondence |
| 8 | Exact selection/ambiguity assurance for integration teams | Preserved item IDs, exact tiny counts, checked existing-solver proposals and explicit unknown negatives | Useful decisions beyond incumbent matching/manual review; accurate business-rule mapping |
These overlap within one platform; do not sum them as independent customers or forecast total revenue from the number of modules. Complete catalog, family directory.
First offer and next decision
The weighted facility release implements the ordinary-plan gap found in public documentation. Eleven generated reports and 7,952 new controls demonstrate supported costs/assignments/capacities. They do not show paid demand or solver superiority. Existing PySAL/OR-Tools capabilities remain the baseline.
Test an assisted review of one permissioned study at an unvalidated AUD 5,000-10,000. At AUD 7,500 and 25 assumed hours at AUD 180/hour, AUD 3,000 remains before other costs; 45 hours exceed the fee in direct labor. Measure material errors, false alarms, preparation/review time, decision changes and support effort before subscription or profitability claims.
Local browser intake now runs around the tested model. Prepare realistic evaluation materials and verify standard-browser export saving before field deployment. Stronger exact bounds and unsupported constraints should follow useful workload evidence. No outreach, purchase, paid compute or customer engagement occurred. The static research catalog and generated reports are published; a hosted customer optimization service has not been deployed.
Retained research decisions
The preceding priorities preserve detailed alternatives, conventional-library comparisons and economics. Direct research backends remain deferred where constants, source proof acceptance, numerical/model bridges or practical performance are unresolved. Family-266 matrix checks are partial certificate evidence; missing local native files prevent reproduction but do not refute its theorem. Three selected family-229 manuscript records now separate graph recovery and two tree companions; source proof acceptance remains incomplete.
Feasibility, 42 utilities, pilot. Counts of ideas, tests or modules are not independent customers or forecast revenue.
Browser release and observed limits: exact on-device review is available, while native in-app saving remains unverified. No additional customer or revenue is inferred.
The finite-field review defers the new backend and keeps a conventional audit integration at low confidence. Existing CAS capabilities and no buyer evidence prevent a standalone-product recommendation. Facility review stays first.
Graph recovery assurance adds finite-baseline calibration, parameter uncertainty and conventional method comparisons. Existing spectral methods already succeed on the selected above-threshold data. The potential value is a recurring review integration, with real-model fit and buyer evidence still absent. It remains below facility review in priority.
Random-SAT benchmark evidence adds exact clause-law controls and distinct solver/censoring evidence. Conventional solvers already handle the selected cases. The source computability claim has no practical efficiency estimate, and threshold-digit commercialization is deferred. This module remains below facility review.
Reaction-network assurance now has finite exact witnesses and conventional comparisons, but a real simulator/import bridge and recurring buyer value remain unvalidated. The source classwise permanence construction has no usable numerical implementation here. Keep this specialist integration below facility review.