Upload your supplier CSV with the messy headers still in it. C2D works out who each supplier is, calculates Scope 1, 2 and 3, and writes the report for you.
Headers never match, countries are free text, and nothing carries a sector. Drop the file in exactly as it comes off your ERP. We match the columns for you.
We try the cheap, checkable things first. The model only ever sees what is left over. When it still cannot tell, the answer is unknown, because we would rather leave you a gap than invent a number.
| Supplier | ISO | Sector | Factor |
|---|---|---|---|
| ••••4821 | MY | Industrial | 0.2447 |
| ••••1096 | SG | Technology | 0.0728 |
| ••••7734 | CN | Consumer Products | 0.4068 |
| ••••2250 | DE | Construction | 0.1121 |
kg CO2e per PPP $ · World Bank EN.GHG.CO2.RT.GDP.PP.KD
industry factors: EPA Supply Chain GHG Emission Factors v1.3
supplier codes masked · rows illustrative
Try it on the sample data, or upload your own export and watch the same four steps run on it.
SWIPE · DRAG · OR USE THE ARROW KEYS
Raw export falling as noise, a scan resolving it left to right, and the ledger it turns into. Same data, same factors, no narration needed.
Then try it on your own fileScope 1, 2 and 3, worked out automatically from the data you already have.
Direct emissions from company‑owned vehicles, fuel combustion, and refrigerants.
Indirect emissions from purchased electricity, heat, or steam.
Supply chain, business travel, commuting, waste, and transportation. Most of a company's real footprint hides here, and messy supplier data is what breaks manual reporting. It is the problem C2D was built to solve first.
Why Carbon2Data
Suppliers are matched against global registries, then given a country and an industry. No manual lookup.
Emission factors come from published reference data. Anything unmapped or low confidence gets flagged, not quietly estimated.
Built against the taxonomies your regulator actually uses, including Bursa Malaysia's CSI framework. Not an invented category list.
Drop in a raw purchase-order or ERP export as it is. Headers and formats get matched for you, so there is no spreadsheet cleanup first.
Producing a figure is the easy half. These are the parts that get you to a filing you can defend.
Send a questionnaire and your supplier answers through a private link. No account, no password, nothing for them to install. Most of them will never sign up for anything, so we stopped asking them to.
Pick what to watch for: a supplier that is still unmapped, one taking a large share of spend, a weak response rate, missing data, a high-risk country. Then pick what happens. An email, a Slack message, a webhook, a task, or an escalation.
Slack, plain webhooks, a connection out to your ESG database, and your own email templates. Nobody needs to live in another dashboard.
Every supplier carries a health score built from how complete its data is, whether it is mapped, whether it answers you, how much of your spend it holds, and the regulatory quality of the country it sits in.
Export to CSV, print to PDF, and read a methodology page that spells out how each figure was produced and which source each factor came from. Your auditor gets to check our working.
Every upload is kept exactly as you sent it, so you can pull back the file behind a number you filed months ago instead of trying to reconstruct it.
Every framework and dataset the engine works against, so you can check the ones that matter to you.
Supply chain emissions, classified by AI and priced with published emission factors.
No report yet
Run the demo sequence or upload supplier data above to generate a classification and emissions report.
We are working with a small number of companies while the product finds its shape. Nothing is held back and there is no card to enter.
Everything below, for as long as this early period runs.
When we do start charging you will hear it from us first, well before it applies to your account, and everything you have already produced stays yours to export.
A supplier or purchase-order export with, at minimum, a supplier name and a spend amount. Country and industry help a lot but are not required, because working those out is most of what the engine does. Column names do not need to match anything.
They are marked unknown and shown to you as unknown. We do not fill the gap with a plausible guess, because a number you cannot trace is worse than a gap you can see. Every estimate also carries a flag for whether it used a real sourced factor or fell back to a global median.
Yes. Everything is calculated from your own spend data first, so you get a complete picture without a single supplier replying. Engagement improves accuracy where suppliers do respond, and the response link needs no account or password, which is the main reason people ignore these requests.
In one place, and only as a fourth resort. We use what your export already says, then legal-entity suffix rules, then the GLEIF registry. Only what survives all three reaches a model, and when the model is not confident the answer stays unknown.
Yes. The platform ships the Bursa Malaysia CSI taxonomy as a default, and any company can upload its own list and map to that instead.
Any time, in CSV or PDF, including the methodology behind each figure. Every file you upload is kept exactly as you sent it, so you can re-download the source behind a number you filed months ago.
For now, yes, and with nothing held back. We are working with a small number of companies while the product settles. If that changes you will hear about it from us before it affects your account.
If you would rather see it run against your own export before signing up for anything, send it over and we will walk through the result with you.