A carbon accounting platform for factories. It does not produce a prettier number; it produces a number you can trace step by step from the result back to the meter reading it came from.
A calculator answers one question: how much is the figure? A record of evidence answers the four questions an auditor, a buyer and a funder ask.
Every value is tied to a source record with a precise identifier: a meter, an invoice, or a documented manual entry - and classified by its evidence grade.
Factor review is an independent human decision, recorded with its owner's name and date, never written from inside the code.
Applicability is a relationship between the factor and its scope - the grid, the period and the boundary - not a checkbox in a factor table.
Evaluation status and publication status are two independent axes. "Not evaluated" is not "refused", and both are valid outcomes.
Most platforms show you the number that passed. Run the gate here and watch a real number get refused - and why.
No gate is relaxed to pass data. A refusal is a complete outcome with a reason and a run ID, not an obstacle in front of another outcome.
Every status carries the ID of the gate run that produced it. A status with no ID raises InvalidStatusOrigin.
NOT_EVALUATED and EVALUATED_REFUSED are never merged into a single display.
Seven links from result to evidence. Open any link to see what it carries - the figure below is recomputed from the factors shown alone.
An immutable snapshot. Any recomputation produces a new snapshot with a new ID; the old one is never replaced.
Engine version and formula ID are part of the trail. A result that cannot name its producer is not traceable, whatever else the trail displays.
The applicability decision is fully independent of factor review. A reviewed, accepted factor can still not apply to this grid or this period - and that is where most refusals happen.
Value and version live in a single record. Calculation, trail, screen and export all read the same record; there is no second copy of the factor inside the interface.
The activity is classified by a governed decision, not guessed from the source name. Scope 2 covers purchased electricity, steam, heat and cooling - not electricity alone.
The observation is reviewed as a draft before it enters the current inventory. A correction keeps the original record standing for the audit trail.
The identifier is enough to locate the evidence without asking a person: source, unit and period. A folder path is not an identifier.
Figures from a synthetic reference case - DERIVED_SYNTHETIC · REFERENCE_ONLY · NOT_PUBLISHABLE. They represent no specific factory and are not published.
These rules are written in code, not in a slide deck.
Acceptance criteria are written before measurement
The gate is not trusted until it is made to fail on purpose
No gate is relaxed to pass data - refusal is the result
Measurements from two environments are never merged into one conclusion
Official ≠ verified ≠ applicable ≠ certified - four independent fields
"Not tested" is never written as "passed"
The number travels with its status - on screen, in export and in print
Every platform shows what it can do. This page also shows what it cannot - because the platform that says so is the only one you can believe about what it can.
| Item | Status | What that means |
|---|---|---|
| Governed engine and trail record | WORKING | Produces figures with a source and a traceable trail |
| Row-level tenant isolation | WORKING | Fully separates every tenant's data |
| Quality gates and governed refusal | WORKING | Stops the error before it reaches the report |
| Factor governance and review | BEING TIGHTENED | A responsible override path, recorded under its owner's name |
| Spinning and weaving adapter | IN PROGRESS | An isolated reference engine and its interfaces |
| Third-party certification | NOT STARTED | We claim no review that has not happened |
Website content last updated: 4 November 2026.
We show the state as it is, without exaggeration, and this page is updated with every advance proven by a test.
We write neither "ready" nor "certified" about our own work, and we issue no assurance opinion. We prepare the file that stands in front of the certification body - and the body is what decides.
One methodology - a record of evidence, a gate and a trail - expands sector by sector. Pick a sector to see its exact position: what is built, what is being discovered, and what is only a target scope.
The methodology also extends to cement, fertilizers, banks, non-banking financial institutions and hospitals - target scopes, not ready modules.
A short film on the role of governance, ISO 14064-1 and IPCC in understanding a carbon footprint figure.
Spinning and weaving demo - a short tour of DEVENUE CFP Smart inside the spinning sector context: from evidence to a traceable number.
A technical visit to the production floor, a survey of existing metering points, and one reference month with a full trail and a real gate. The expansion decision is then built on what the month produced, not on a promise in a slide.