EmberProof — CBAM emissions packet desk for Indian metal exporters

1. One-liner

A guided desk that lets an Indian steel or aluminium exporter collect upstream supplier emissions, compute EU-method embedded carbon, and hand their EU buyer a ready-to-submit CBAM packet.

2. Trend signal — why now?

CBAM’s definitive (paying) phase went live 1 January 2026. There is no SME exemption — every shipment of CBAM goods into the EU now carries a carbon cost, and the EU buyer (the “authorised declarant”) either gets verified actual emissions from the exporter or applies punitive default values.

What’s actually happening to Indian exporters right now, not in theory:

• Indian MSME steel exporters are reporting shipment seizures, order cancellations and held consignments since the payment phase began. At least 10 Indian consignments were held at European ports in January 2026 over missing or incomplete carbon declarations. (Business Standard, Jan 2026)
• A Mumbai exporter lost a 7,000-tonne steel order after CBAM added ₹5–6 crore to costs, killing the deal. (KNN India)
• Carbon levies for many MSMEs have tripled to €240–300 per tonne because they’re being pushed into default values instead of supplying actual data. Using real, lower data can save 15–25% of CBAM cost. (KNN India)
• The EU fixed the Annex IV CBAM Communication Template and the Q1-2026 default-value list in December 2025 — so the target output of a tool is now a stable, known spec, not a moving goalpost. (Carboneer)

The reason data isn’t flowing: the exporter has to chase dozens of small upstream domestic suppliers (re-rolling mills, power, scrap) for installation-level numbers in EU MRR methodology — and most of those suppliers have zero ESG infrastructure. “Even if one supplier in the chain fails to give compliance data, MSMEs are pushed into punitive default values. No small exporter has the margins to absorb that kind of shock.” (Business Standard, Jan 2026)

Provenance:

• Signal 1 (Demand): Indian MSME steel exporters hit by port seizures, cancelled 7,000t order (₹5–6 cr added), levies tripled to €240–300/t; 3,000–4,000 direct + 25,000–30,000 indirect exporters exposed — https://knnindia.co.in/news/newsdetails/msme/indian-steel-msme-exporters-hit-as-eus-cbam-payment-phase-begins & https://www.business-standard.com/industry/news/indian-msme-exporters-hit-by-seizures-as-cbam-payment-phase-begins-126012801235_1.html — Jan 2026
• Signal 2 (Feasibility): EU fixed the Annex IV CBAM Communication Template + Q1-2026 default values in Dec 2025; MRR embedded-emissions math is now a deterministic, implementable spec; cheap LLM extraction handles messy supplier inputs (bills, fuel logs, scrap mix) — https://carboneer.earth/en/2025/12/cbam-definitive-period-scope-extension/ — Dec 2025
• Signal 3 (Economic): CBAM projected to raise €80B/yr by 2040; verified actual data saves 15–25% of CBAM cost per shipment; funded software (CarbonChain, Coolset) is flowing — but to the EU importer end, not the supplier end — https://www.spglobal.com/sustainable1/en/insights/special-editorial/eu-carbon-border-adjustment-mechanism-to-raise-80b-per-year-by-2040 — 2026 Category: Regulatory arbitrage

3. The opportunity

The CBAM software market exists — but it’s built for the wrong end of the chain. CarbonChain, Coolset, Zero Circle, EnCarbonSys, and the Big-Four consultancies all sell to the EU importer / authorised declarant or to mid-market/enterprise. Their job is to help the buyer calculate liability and buy certificates.

Nobody affordable is sitting on the supplier side, holding the hand of a ₹20–80 crore-turnover Indian metal exporter who just got an email from their German buyer that says: “Send me your installation-level emissions in the EU Summary Communication format or I price you at default values.” That exporter has:

• No idea what an “installation” or “MRR methodology” is
• A pile of upstream suppliers who’ve never measured a thing
• A real deadline, because the buyer is already factoring carbon intensity into supplier selection
• A direct 15–25% price gap between supplying data and not supplying it — i.e., the difference between keeping the order and losing it to a competitor who can

The incumbents are too expensive and too importer-shaped for this customer. The wedge is a ₹-priced, exporter-shaped, guided desk that turns “the buyer wants my carbon data” from a panic into a 2-hour workflow that ends in a buyer-ready Excel packet.

This is distinct from generic “carbon accounting.” It does exactly one job: produce the CBAM Communication Template the EU buyer will accept, with the lowest defensible embedded-emissions number, fast.

4. Target market

• Primary customer: Owner / export manager / compliance lead at an Indian steel, aluminium, or steel-product exporter doing direct or Tier-1 trade with EU buyers — roughly ₹20–200 crore annual turnover, 20–300 staff, exporting re-rolled steel, fasteners, aluminium sections, castings, fittings.
• Why they buy: “My buyer will pay me 15–22% less or drop me entirely if I can’t hand over CBAM-compliant emissions data, and I have no clue how to produce it.” It’s a revenue-loss problem disguised as a compliance problem — the strongest kind.
• Rough TAM reasoning: 3,000–4,000 direct Indian exporters + 25,000–30,000 indirectly exposed MSMEs (CleanCarbon estimate). Even capturing 1,000 direct exporters at a mid-tier price is a clean $1–2M ARR business. Add Turkey, Vietnam, and other CBAM-exposed metal exporters and the same engine extends.
• Why now for them: The paying phase started Jan 2026. The first real CBAM cost has already hit invoices, shipments are already being held, and EU buyers are already asking for the data. This is a this-quarter fire, not a someday concern.

5. Product sketch (MVP)

• Buyer-request intake: paste or forward the EU buyer’s data request; the desk identifies exactly which CBAM goods, CN codes, and installations are in scope.
• Supplier chase-and-collect: generates a plain-language (Hindi/regional + English) data request to each upstream supplier with a dead-simple form — “upload last quarter’s electricity bill, fuel purchase, production tonnage” — no jargon.
• Messy-input extraction: reads supplier bills, fuel logs, and scrap-mix sheets (PDF/photo/Excel) and pulls the numbers needed for the MRR calculation.
• Embedded-emissions engine: computes direct (Scope 1) + indirect (Scope 2) embedded emissions per product using EU methodology, with the grid/fuel emission factors baked in.
• Default-vs-actual gap meter: shows the exporter, in euros per tonne, how much they save by submitting actual data vs taking the default — the “why bother” number, up front.
• Buyer-ready packet export: outputs the exact EU Annex IV Summary Communication Excel tab, plus a clean PDF cover the buyer’s declarant can accept.
• Audit trail: keeps the source documents and calculation behind every number, so when a verifier or buyer questions it, the evidence is one click away.

6. AI angle — what’s load-bearing

Two places AI does real work, not decoration:

1. Extraction from chaos. Upstream Indian suppliers will never submit clean structured data. They’ll send a photographed electricity bill, a WhatsApp of a fuel ledger, a scrap-mix note. LLM + vision extraction turns that into the structured inputs the MRR calculation needs. Without it, this is a consulting engagement, not software.
2. Methodology translation. The desk reads the buyer’s specific request and the relevant CBAM guidance and tells this exporter, in their language, exactly which inputs map to which template fields — collapsing what a CBAM consultant charges lakhs to do into a guided flow.

Remove the AI and you’re left with a blank Excel template and a PDF of EU regulations — which is exactly the unusable status quo the exporter is drowning in. The AI is the difference between “here’s the form, good luck” and “answer these 8 questions and we’ll generate the form.”

7. Localization angle

This is the localization play — that’s the whole moat against CarbonChain/Coolset.

• Language: supplier-chase messages and the exporter’s guided flow in Hindi + regional languages; EU output stays in the EU’s English template.
• Pricing: a ₹-denominated tier (think ₹4,000–15,000/mo) that works for a ₹40 crore exporter where a €500+/mo enterprise tool doesn’t.
• Local data realities: Indian grid emission factors, coal-heavy re-rolling furnaces, scrap-based EAF routes baked into the engine — the exact inputs a generic global tool fumbles.
• Local distribution: EEPC India, ISSDA, AIIFA (re-rolling mills), FIEO, and regional steel/aluminium export clusters (Ludhiana, Rajkot, Mumbai) are dense, listed, and reachable.

A generic global product will always be importer-shaped and euro-priced. The India-exporter cut wins on price, language, and the specific upstream-supplier-chasing workflow that’s India’s actual bottleneck.

8. Business model — path to $1M–$5M ARR

• Pricing: tiered SaaS. Starter ₹4,000/mo (single product line, a few suppliers); Growth ₹12,000/mo (multiple CN codes, supplier portal, audit trail); plus a per-packet option (₹6,000–10,000/packet) for occasional exporters. Optional verifier-handoff add-on.
• ACV: ~₹1.0–1.4 lakh ($1,200–1,700) blended on subscription; higher where per-shipment volume is high.
• Math to $1M ARR: ~700 exporters × ₹1.2 lakh/yr ≈ ₹8.4 cr ≈ $1M ARR. That’s <20% of the 3,000–4,000 direct-exposed pool.
• Math to $5M ARR: ~2,500–3,000 paying exporters across India + Turkey + Vietnam, plus a higher-ACV “managed packet” service for the largest, and a verifier-marketplace cut. Requires geographic expansion and proving retention past the first compliance cycle.
• Expansion path: ACV grows as scope widens — more CN codes, more suppliers in the portal, then adjacent CBAM goods (cement, fertiliser), then the verification handoff, then quarterly recurring packets as buyers demand fresh data each period.

9. Go-to-market wedge — first 100 customers

• Scrape the export bodies. EEPC India, ISSDA, AIIFA and FIEO publish member directories of metal exporters. Pull the 800–1,200 with EU trade exposure; email/WhatsApp a personalised one-pager: “Your EU buyer is asking for CBAM emissions data. Here’s your packet in 2 hours, or you eat €240–300/tonne in default values.”
• Ride the panic press. The Business Standard / KNN / ET coverage of seizures and lost orders has primed the market. Run targeted LinkedIn + WhatsApp outreach to export managers reacting to those stories, with a free “default-vs-actual gap calculator” as the hook.
• Partner with CHAs and export consultants. Customs House Agents and freight forwarders already handle these exporters’ EU shipments and are hearing the CBAM complaints first-hand. Revenue-share referral; they bundle EmberProof into the export paperwork they already manage.
• Cluster workshops. Run a free “How to answer your buyer’s CBAM request” webinar through one regional steel cluster association (Ludhiana / Rajkot). Convert attendees who already have a live buyer request — these are pre-qualified, deadline-driven buyers.

Conversion math: a 1,000-exporter scrape, 5% reply, 30% of replies are actively panicking with a live buyer request, half close in the first cycle ≈ 7–8 paying from one list — repeat across 3–4 bodies + referral channel to clear 100 in two quarters.

10. Build complexity — justification

Medium. The hard parts are (a) implementing the EU MRR embedded-emissions calculation correctly with current emission factors, and (b) robust extraction from messy supplier documents — both off-the-shelf-API-doable but requiring care and domain validation. The output template is a fixed, published EU spec. No custom models, no hardware, no marketplace chicken-and-egg. A 2–3 person team (one with CBAM/carbon-accounting domain knowledge) ships a credible v1 in ~3–4 months. The methodology is the risk, not the stack.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Helping exporters comply with an EU regulation — squarely legal.
Ethical — no harm / dark patterns	✅	Reduces penalty exposure with honest, auditable data. No greenwashing — actuals only.
Market exists (evidence above)	✅	Live seizures, lost orders, tripled levies, dated Jan 2026.
1–5 person team can build this	✅	2–3 people, ~3–4 months.
Launchable with <$50K / ₹40L	✅	Off-the-shelf APIs, no capex.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	17/20	Hair-on-fire: real orders lost, shipments seized, €240–300/t penalty this quarter.
Demand evidence	15	13/15	Multiple independent 2026 sources, hard numbers, named exposed pool of thousands.
Build feasibility	15	10/15	Methodology + messy-extraction are non-trivial; output spec is fixed. ~3–4 months.
Distribution clarity	15	11/15	Named, listed export bodies + CHA referral + primed press. Conversion still to prove.
Revenue mechanics	15	11/15	₹-pricing works for wallet; $1M needs ~700 of 3,000–4,000 — credible but real selling.
Time to first revenue	10	7/10	Live deadlines pull buyers fast; first paid in ~6–10 weeks plausible.
Defensibility	10	5/10	Moat is the India-supplier workflow + audit data + body relationships, not the math. Copyable.
Total	100	74/100

13. Qualitative modifiers

Founder-fit tags

domain-expertise-required · sales-heavy — needs someone who genuinely understands CBAM/MRR methodology and can sell into Indian export clusters and associations.

Key assumptions to validate (3–5)

1. Assumption: Exporters will pay ₹4k–15k/mo to a software desk rather than hand the whole thing to a consultant. How to test: 30 calls with EEPC/ISSDA member exporters who have a live buyer request — offer paid pilot, measure close rate.
2. Assumption: Upstream suppliers will actually return usable data to a simple form (even photos of bills). How to test: run the chase flow with 5 real exporters’ supplier lists; measure response rate and how often extraction yields a defensible number.
3. Assumption: EU buyers accept a packet generated this way without bouncing it back. How to test: get 3 pilot exporters’ packets in front of their actual EU declarants and confirm acceptance.
4. Assumption: The actual-vs-default savings are large enough that the tool pays for itself in one shipment. How to test: compute real gap on 10 live shipments; confirm it exceeds annual subscription.

Risk flags

1. Regulatory risk: CBAM rules, default values, and templates evolve; the engine must track EU changes each cycle. Mitigated by the spec being newly fixed (Dec 2025) but ongoing maintenance is real.
2. Verification dependency: From 2026, actual emissions ultimately need ISO 14065 / EU-recognised verifier sign-off. The desk produces the data; it can’t replace the verifier — must partner, not over-promise.
3. Incumbent down-market move: A funded importer-side player (CarbonChain) could ship a cheap supplier module. Counter with India-specific workflow, language, ₹-pricing, and association lock-in before they bother.
4. Trade-policy whiplash: India–EU trade talks or CBAM carve-outs could soften the penalty and cool urgency. Watch the bilateral negotiations.

14. Structured verdict

Score:                  74/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Carbon-accounting/CBAM domain expert paired with a sales-led co-founder wired into Indian export clusters
Time to revenue:        6–10 weeks (live buyer deadlines pull customers fast)
Capital to launch:      ₹6–12 lakh ($7–14K)
Top 3 assumptions to validate first:
  1. Exporters pay for software vs. full consultant — 30 cluster interviews, measure paid-pilot close rate
  2. Upstream suppliers return usable data to a simple form — run the chase flow on 5 real supplier lists
  3. EU declarants accept the generated packet — put 3 pilot packets in front of real EU buyers
Kill criteria:
  - Abandon if <15% of 40 cold-outreach exporters with live buyer requests show willingness to pay within 3 months
  - Abandon if supplier data-return rate stays below ~40%, forcing default values anyway (product can't deliver its core promise)
  - Abandon if an India–EU CBAM carve-out removes the penalty for Indian metal before launch

15. Next step — 1-week validation sprint

• Day 1–2: Pull 60 EU-exposed metal exporters from EEPC India / ISSDA member lists. Build the one-page “default-vs-actual gap” pitch.
• Day 3–4: Personalised WhatsApp/email + calls. Target 15 live conversations; for each, ask: do you have a buyer asking for CBAM data right now, what are you doing about it, would you pay ₹X/mo for a packet in 2 hours? Run one real packet by hand for the most willing exporter to test the supplier-chase + extraction loop end-to-end.
• Day 5: Decide go / no-go. Go if ≥6 of 15 have a live buyer request AND ≥4 say they’d pay for the packet AND the hand-run packet produced a defensible actual-emissions number lower than the default. Otherwise iterate the wedge or shelve.

A falsifiable result: either exporters with live deadlines commit to paying and the manual packet beats the default value — or they don’t, and this stays a consultant’s job.