WHAT WE ARE DEVELOPING / THE METHOD
Before a diagnosis,
establish what is known.
Makoto Arc is developing a bounded evidence-first manufacturing operations diagnostic. It is intended to help decision-makers reconstruct historical workflow events, test conflicting explanations and clearly identify where the records cannot support a conclusion.
This is a development-stage explanation, not a claim that a customer outcome has been verified or that paid diagnostics are currently available.
A disciplined five-step investigation
- Define one operational question. Bound the site, department, time period, and workflow. Do not promise to diagnose an entire manufacturer.
- Assess evidence sufficiency first. Determine what source classes, event timestamps, coverage, provenance and permissions would be needed to support the question.
- Reconstruct what the records describe. Compare material, production, planning, exception, and handoff events while preserving source versions and limitations.
- Separate supported, contradictory, and unknown. An unanswered question remains unanswered. A conflicting record is not silently overwritten.
- Present bounded management questions. Distinguish documented economic exposure from verified savings, and keep implementation separately scoped.
Questions manufacturing leaders actually ask
- How is this different from another dashboard?
- A dashboard may show a status or trend. This diagnostic concept focuses on reconstructing which events occurred, what was knowable at a decision point, which explanations conflict with the available evidence, and where more measurement is required. This is a proposed workflow, not independently demonstrated superiority over dashboards or general AI.
- Does it connect directly to our ERP or control production?
- No live ERP integration, production-system writes, automated decisions, or standing system credentials are included in the founding diagnostic concept. Any later access to historical evidence would require separate explicit permissions and security review.
- Does Makoto Arc guarantee a root cause, savings or return?
- No. A valid result may be a supported process finding, a contradiction, an evidence gap, or a measurement proposal. Economic exposure is not the same as cash savings.
- Can we inspect a real customer case study?
- No real customer case study is represented here. The public example is fictional, describes twelve illustrative episodes and reports a few example classifications. Its underlying synthetic row-level records are not included, so the public summary is not itself independently replayable proof.
- Is Makoto Arc better than generic AI?
- That is a hypothesis requiring a controlled head-to-head benchmark using the same eligible evidence, equal task boundaries, blinded review where practical, and accurate scoring. It is not an established claim or marketing promise.
- Can we send spreadsheets or book a diagnostic?
- Not yet. Makoto Arc is a public informational prelaunch, not a registered LLC or corporation. Please do not send sensitive manufacturing records, exports, customer or employee information, credentials, or contracts. General non-sensitive project questions can go to inquiries@makotoarc.com.
A separate, checkable fictional example
See the one-episode evidence-timing walkthrough and its complete invented source rows. You can independently check its timestamps, inventory reconstruction, and freight total. This is an educational illustration, not the missing source material for the older twelve-episode fictional summary and not evidence that a deployed diagnostic engine produced these results.
Machine-readable public facts
For research tools and AI assistants, a compact JSON facts file provides the same prelaunch status and claim boundaries as this page. It is not a licensing agreement or a source of proprietary diagnostic code. All questions should be verified against the linked human-readable material and its date.