Every figure this platform produces can be traced back to the data that originated it.
Environmental, climate, and regulatory reporting management for industrial and service organizations. Traceability isn't a declared policy — it's built into the data model.
Each figure keeps the measurement that produced it, the factor and formula used to compute it, and the validation level that supports it. When that level doesn't exist, the platform declares it unvalidated rather than fabricating the figure.
Auditability isn't declared in a document. It follows from how the data is stored.
An environmental platform can promise traceability in its brochure and be unable to sustain it in its database. EcoCyclus inverts the order: traceability is a consequence of three construction decisions — how the measurement is stored, how formulas are versioned, and how the validation level is resolved at the moment someone reads the figure.
That construction has an uncomfortable consequence the platform accepts: the validation level is not assumed, it's declared — and today most of the portfolio is declared unvalidated. That "unvalidated" is the argument, not the debt: a system that can't say so can't prove otherwise either.
(source × geography)
Model validation is per geography: calibrating one region doesn't validate another. EcoCyclus is not "audited" or "certified" as a platform — the assurance level is an axis the issuer's data declares, not a state the platform holds.
In adaptation, the standard itself declares there are no core indicators. Only illustrative ones.
It's the one front where arriving today isn't arriving late.
For an issuer or holder of a green bond, EcoCyclus operates the project from its own screen: the fund allocation ledger keeps its complete history, the asset is linked to the declared adaptation measure, and each result is issued under a named regime with its version.
Mitigation is known territory: established indicators and heavy competition. Adaptation isn't — the standard admits it has no core indicators there. The platform reaches that gap with the ex ante counterfactual declared against a "no action" scenario, with the protected arm derived from the model rather than typed in, and with ICMA GBP and EuGBS coexisting without one overriding the other.
The adaptation counterfactual does not model defense failure, overtopping, or fragility. Regarding EU Taxonomy, only eligibility is resolved: not alignment, not DNSH. EcoCyclus does not certify green bonds — it produces the document; certification is issued by an external reviewer.
Risk is calculated on the asset's coordinate, not on the country's average.
A national average doesn't distinguish two plants twenty kilometers apart across a watershed. In geographies where aggregated data is coarse — LATAM among them — that difference decides the risk reading.
Below that resolution runs a probabilistic engine: exceedance curve, AAL, and PML, with inter-model uncertainty propagated to the result rather than averaged out beforehand. Coastal loss is computed by coordinate, with a declared uncertainty band, over RCP scenarios.
Not all hazards reach the same point of the method, and the platform distinguishes this on screen.
Fluvial flood, coastal flood, earthquake, and wind. They produce AAL, PML, and asset rating.
Extreme precipitation, water stress, and drought. They are reported as exposure and trend. For extreme precipitation the damage function is missing; for water stress and drought computing one doesn't apply: they are composite indices, not physical intensities, and turning them into loss would count vulnerability twice. The platform declares this rather than filling the gap.
The two flood families. Earthquake and wind are computed, but not contrasted against a "no action" scenario.
Transition risk runs on NGFS scenarios, and physical and transition are not summed into a single number: they don't share a measurement space, and collapsing them would produce a figure neither of them sustains.
On the same base, a supplier risk screening runs, with the scope declared below.
Climate resilience screening for LEED v5 (IPp1): the same risk engine issues the assessment report in the format of USGBC's official template, asset by asset.
Earthquake outside the U.S.: hazard well-founded, loss indicative — screening level. Supplier risk: screening-grade, not actuarial. Transition risk: not aggregated with physical risk. The geographic domain is a declared set of countries, not global coverage. LEED IPp1 requires assessing twelve hazards; the engine gives solid coverage of flooding and partial coverage of water stress and coastal risk — the rest is being extended.
Scope 1, 2, and 3 under two standards, and the trace open down to the measurement.
The same inventory is issued under GHG Protocol or under ISO 14064-1, switching between taxonomies without manual recomputation. Scope 2 is calculated by both methods — location-based and market-based — and both figures coexist, because depending on the framework it may be appropriate to report either.
Each indicator opens down to the computation and the measurement that produced it, with the formula and factor visible in the trace.
The emission factor database is searchable by multiple fields at once, keeps the source each factor came from, converts gases to CO₂e with the GWP version it declares, normalizes units, and resolves the factor that corresponds to each geography. Regulatory thresholds are evaluated on the inventory itself, not on a copy.
The engine is deterministic and auditable. It's standard capability in the sector, and the platform covers it without inflating it.
The computation is not automated with AI: the engine is deterministic, and that is the argument.
Footprints that circulate between organizations without inventories touching each other.
Cradle-to-gate product carbon footprint under ISO 14067, with the CFP · Clause 7 report and allocation method reconciled with the standard.
Reconciliation with the corporate inventory is by construction, not by post-hoc reconciliation: the engine runs on declared inputs, and the separation between fossil and biogenic carbon is enforced by the data model, not by convention.
On top of that, there is a network of footprints between organizations: one publishes, another consumes with precedence resolution. When the supplier publishes a new version, whoever was consuming the previous one is notified — obsolescence detects itself.
The network does not share data between organizations: it publishes already resolved footprints. The boundary between organizations is respected in the backend, not by convention.
Project footprint for LEED v5 (MRp2): the same cradle-to-gate engine computes embodied carbon for construction materials by location, drawing on an EN 15804 catalog (ÖKOBAUDAT) and DEFRA.
Cradle-to-gate and climate change only: not cradle-to-grave, not multi-impact, not full life cycle. LEED MRp2: industry-average grade, not manufacturer-verified EPD — that scoring tier is covered alongside a specialist partner.
Coverage across several frameworks at once, without imposing hierarchy between them.
The coverage engine evaluates the report against several frameworks simultaneously, and ESRS-GRI equivalences live inside the model, not in a side spreadsheet.
Qualitative disclosure responses are a first-class entity: they are versioned and reused across frameworks, rather than living loose inside a document.
Attribution to the SDGs runs over the official GRI-SDG crosswalk. The indicator catalog is navigated by ESRS topic. EU Taxonomy eligibility screening is resolved over NACE.
There are two things the platform does not emit, and these are not omissions but properties of the discipline: it does not declare alignment with EU Taxonomy, and it does not emit an SDG compliance percentage. The second is unemittable by construction, deliberately — SDG indicators are defined at country level, not at company level.
EU Taxonomy: only eligibility, not alignment, not DNSH. SDG attribution: does not emit a compliance percentage.
Risk categories, criteria, and review cycles, and the target gap grounded in real water risk.
The risk module has categories, evaluation criteria, and review cycles as model entities — ISO 31000 nomenclature — with a matrix and per-entry declaration.
On that base, the SBTi target gap analysis is computed against the water risk of the per-asset engine, with WRI Aqueduct underneath. It's the same water stress engine as the risk section and the same water metric the inventory reports under GRI 303: the gap doesn't live in a separate spreadsheet.
It is gap analysis, not target validation: SBTi issues validation, not the platform.
If the method is the one your organization needs to be able to show, write.
EcoCyclus speaks with three types of interlocutor, and each enters through a different door:
Through the green bonds and adaptation section.
Through inventory, product footprint, and disclosure.
Through physical and transition risk, asset by asset.
Who is accountable for this
EcoCyclus is built by Fernando Amar, based in the Province of Buenos Aires, Argentina.
The method comes from practice: sustainability with audit reflex. From there follows the decision that orders everything else — when the data doesn't sustain a figure, the platform declares what's missing rather than filling the gap.
And the geography isn't a contact detail: it's where the national average is most misleading, and where the question that originated the platform has a different answer depending on the coordinate.