ShipClear — export clearance gate for India’s spice exporters

1. One-liner

Screens every food shipment against the destination’s live rules and flags what gets it rejected — before the vessel sails.

2. Trend signal — why now?

The bite is getting sharper every quarter, and it’s measurable.

• RASFF alerts on Indian spices are climbing and India is the #1 flagged origin. 279 spice/herb issues were logged in the EU’s RASFF in 2025 (up from 275 in 2024), and India was the single most-notified origin country for herbs and spices at 23.6% of all notifications — ahead of Brazil, Thailand, Turkey, China. The leading cause was exceeding pesticide Maximum Residue Limits (MRLs), then plant toxins, then mycotoxins. (ScienceDirect RASFF herbs/spices analysis)
• The EU keeps ratcheting up “reinforced controls” on specific Indian commodities, and the list moves. Cumin (HS 0909 31) from India has been under reinforced testing at 30% physical-check frequency since January 2025; peppers and Capsicum/Pimenta spices are under higher scrutiny for ethylene oxide. The reinforced-control annex is updated by EU Implementing Regulations (e.g. 2026/194) — exporters are told to “stay regularly updated” but have no tooling that tells them when their commodity’s status changes. (AGRINFO Jan 2026 update, CBI buyer requirements)
• The losses are large, real, and land on small exporters. A Mumbai exporter lost a $500,000 cumin contract to Germany when chlorpyrifos was detected at 0.02 mg/kg; a Gujarat curry-powder maker took a ₹1.2 crore hit (recall + cancelled 3-year contract + blacklisting) over 0.08 mg/kg ETO; 2024 saw 1,200+ India spice alerts costing ~$200M in lost shipments. A single bad batch can trigger automatic rejection and blacklisting. (Sadbhaav rejection breakdown, foodtechnetwork rejections 2020–26)

Provenance:
  - Signal 1 (Demand): India = #1 RASFF-notified origin for herbs/spices (23.6%), 279 alerts in 2025, MRL breaches the top cause — https://www.sciencedirect.com/science/article/abs/pii/S0956713525006048 — 2025
  - Signal 2 (Feasibility): EU reinforced-control status is public but moves via Implementing Regulations; cumin at 30% check since Jan 2025 — destination MRL/rules data is ingestible and updatable by an AI pipeline — https://agrinfo.eu/book-of-reports/temporary-increased-official-controls-on-foods-from-certain-countries-january-2026-update/ — Jan 2026
  - Signal 3 (Economic): 42,740 active APEDA exporters, $28.59B APEDA exports (55% of agri exports); ~$200M lost India spice shipments in 2024; six/seven-figure single-contract losses documented — https://apeda.gov.in/about-us & https://sadbhaavspices.com/top-5-spice-export-rejection-reasons/ — FY24-25
  Category: Regulatory arbitrage

3. The opportunity

There are two kinds of software in this market and a gap between them.

On one side: export documentation generators — OptiExim, Visualexportems (2,500+ companies), ExportDoc, Pazago, Hexponent. They turn one data entry into a commercial invoice, packing list, certificate of origin, e-way bill. Useful, crowded, commoditised. None of them tells you whether the shipment will be rejected at the other end. They format your paperwork; they don’t judge it against the importing country’s rules.

On the other side: NABL labs and compliance consultants — they run the chemistry, charge per sample, and are out of reach for the long tail of small exporters (only ~30% of exporters have in-house labs). They tell you what’s in the product, days later, after you’ve already committed the lot.

The gap: a pre-shipment clearance gate that, the moment an exporter is about to ship, cross-checks the consignment — commodity + HS code + destination + the lab CoA they already have + the documents — against the destination country’s live rules: current MRL list for that commodity, reinforced-control/increased-check status, banned-substance flags (ETO, specific pesticides), required additional declarations, correct botanical naming, certificate validity and timing. It returns a plain-language GO / FIX / HOLD verdict with the exact defect, in the 24–72 hours before the vessel cutoff when it can still be fixed.

This is not the chemistry. It’s the avoidable layer on top of it — the declaration errors, the destination-rule mismatches, the “we shipped to the EU using the US MRL assumption” mistakes, the “this commodity went to 30% reinforced check last month and nobody told us” misses. Those are exactly the failures that cause missed sailings, demurrage, documentary RASFF rejections, and blacklisting — and they’re catchable with software.

4. Target market

• Primary customer: Owner / export manager at a small-to-mid Indian spice & processed-food exporter — 5–80 staff, ₹5–150 crore turnover, shipping whole/ground spices, blends, rice, processed foods to the EU, UK, US, Gulf, SE Asia. Holds APEDA RCMC, Spices Board CRES, FSSAI; ships dozens to a few hundred consignments a year.
• Why they buy (their words, paraphrased from trade sources): “Each country has different MRLs and the rules keep changing — by the time we find out a limit moved, the container is already gone.” A single documentary or MRL-mismatch rejection wipes out a quarter of margin, and a blacklisting can end a buyer relationship worth crores. (CBI, Sadbhaav challenges)
• Rough TAM reasoning: 42,740 active APEDA registrations; even if only the spice/processed-food long tail of ~8,000–15,000 SME exporters is addressable, at ₹2,000–8,000/mo that’s a ₹20–140 crore ($2.5–17M) annual ceiling for India alone — before the same gate ports to Vietnam pepper, Sri Lanka tea, Turkey/Brazil origins facing identical RASFF pressure.
• Why now for them: EU reinforced testing on cumin hit 30% in Jan 2025; APEDA flagged stricter EU import controls in April 2026; the India-EU trade-pact momentum means more volume into the most rules-heavy market, not less. The cost of a mistake just went up and the rules just got more fluid.

5. Product sketch (MVP)

• Shipment screen: enter commodity + HS code + destination + quantity; upload the CoA / lab report and the draft invoice & packing list. Get a GO / FIX / HOLD verdict in minutes.
• Destination rule-check: does this commodity’s measured residues clear the destination’s current MRLs? Is it on a reinforced-control / increased-physical-check list right now? Any banned-substance flags (ETO, specific pesticides) for that lane?
• Document defect catcher: wrong/absent botanical (scientific) name, missing additional declarations the destination requires, certificate-of-origin / phyto / health-cert validity and timing vs vessel cutoff, HS-code mismatch.
• “Status changed” alerts: watch the commodities + lanes an exporter actually ships and ping them when a destination MRL, reinforced-control percentage, or banned-list entry changes — the thing they currently miss until a container is rejected.
• Plain-language defect report: each flag explains what’s wrong, why it would get rejected, and the fix — written for an export manager, not a chemist.
• Audit trail: a timestamped record of every pre-shipment screen, exportable for the buyer’s due-diligence file (importers increasingly demand this).
• Multi-destination compare: same lot, three buyers — which markets it can legally clear today.

6. AI angle — what’s load-bearing

Remove the AI and this doesn’t exist as a small-team product. Three jobs are AI-load-bearing:

1. Ingesting and structuring fluid regulation. Destination MRL tables, EU Implementing Regulations, reinforced-control annexes, banned-substance lists — published as dense legal PDFs and tables across dozens of jurisdictions, updated unpredictably. An LLM pipeline reads, diffs, and normalises these into a queryable rules layer; a human team could not keep 30+ lanes current by hand.
2. Reading the exporter’s messy inputs. CoAs, lab reports, draft invoices and packing lists arrive as scanned PDFs and inconsistent formats. The model extracts measured residues, commodity descriptors, botanical names, quantities and certificate dates and maps them onto the rules layer.
3. Judging + explaining. Matching a measured residue against the right MRL for the right commodity-destination pair, spotting the missing additional declaration, and writing the fix in plain English is a reasoning task, not a form. That judgement is the product.

7. Localization angle

India-first by design, and the wedge is the geography. The buyer set is Indian (APEDA / Spices Board / FSSAI registered), the distribution channels are Indian (exporter directories, Spices Board clusters, CHA/freight-forwarder networks, WhatsApp), pricing is rupee-micro (₹2,000–8,000/mo where a $200/mo global tool won’t land), and the originating-country quirks (Indian botanical naming conventions, ICEGATE/eSANCHIT document flow, the specific commodities India gets flagged on) are baked in. A generic global “trade compliance” suite (PwC Navigate, enterprise GTC tools) is priced and built for large importers, not a ₹40-crore Kerala spice house. The same shape ports later to Vietnam (pepper), Sri Lanka (tea), and other RASFF-pressured origins — but India spices is the beachhead because the pain is hottest and the buyer list is a public directory.

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

• Pricing: tiered SaaS. ₹2,499/mo (Starter, up to ~25 screens/mo, 1 destination region), ₹5,999/mo (Growth, ~100 screens, all destinations + status alerts), ₹12,999/mo (Pro, unlimited + multi-user + audit export + API for the forwarder). Optional per-screen overage.
• ACV: blended ₹70,000 ($840) landing toward ₹100K+ as exporters move to Growth/Pro.
• Rough math to $1M ARR: ~1,000 paying exporters × ~₹7,000/mo blended ≈ ₹8.4 crore ≈ $1M ARR. Out of an 8,000–15,000 SME-exporter pool, that’s 7–12% penetration — achievable for a focused vertical tool with a directory-driven funnel.
• Rough math to $5M ARR: ~3,500 India exporters on higher tiers plus the port to one or two other origin countries (Vietnam, Sri Lanka) and an upsell of the audit-trail/due-diligence export that importers will pay for. Requires the rules layer to be trusted enough that a rejection-after-GO is rare.
• Expansion path: more lanes per account → status-alert seats → buyer-facing due-diligence exports → forwarder/CHA API integration (they screen on the exporter’s behalf and bundle the cost).

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

• Scrape the public APEDA / Spices Board exporter directories (commodity + state + contact are listed), filter to spice & processed-food SMEs in the top clusters — Kochi/Kerala, Gujarat (Unjha cumin), Rajasthan, Andhra (chilli), Mumbai. Send each a free “rejection-risk screen” of their actual last shipment: “Here are 3 things on your cumin-to-Germany lane that would fail an EU check today.” A personalised teardown, not a pitch.
• Go where the loss stories live: the Federation of Indian Spice Stakeholders, Spices Board CRES clusters, and commodity associations are actively alarmed (FISS publicly fears a 40% export drop). Run a free webinar — “the 7 documentary errors behind India’s RASFF rejections” — co-hosted with a known export consultant; convert attendees to trials.
• CHAs and freight forwarders as a channel: they feel the demurrage and the angry-client call when a container is held. Offer them a co-branded screen they run for clients; rev-share. One forwarder can funnel dozens of exporters.
• Hijack the rejection moment: RASFF notifications and many border rejections are public. When an Indian exporter gets flagged, that company and its competitors on the same commodity are the warmest possible leads — a “this just happened on your lane; here’s how to not be next” outreach converts.
• Beachhead one commodity: own cumin-to-EU first (highest alert rate, under 30% reinforced check). Be the obvious tool for that lane, then widen.

10. Build complexity — justification

Medium. The app shell (upload, screen, verdict, alerts, audit log) is a standard web stack with off-the-shelf LLM APIs for extraction and reasoning — a pair could ship that in 8–10 weeks. The hard, ongoing work is the rules layer: ingesting and keeping current the destination MRL tables, reinforced-control annexes and banned-substance lists for the launch lanes (start with EU + UK + Gulf + US for the top 5 commodities, not all 30 lanes at once). That’s a curated AI-assisted data pipeline plus a domain expert who validates it — buildable by 1–3 people in ~4 months if scoped to a beachhead, but it never “finishes,” which is also the moat.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Compliance-assist tool; no licence needed. Must disclaim it is not a lab/legal guarantee.
Ethical — no harm / dark patterns	✅	Reduces rejections and food-safety incidents; net-positive. Must avoid false “PASS” assurance — framed as risk-flagging.
Market exists (evidence above)	✅	42,740 APEDA exporters, $200M documented losses, India = #1 RASFF origin.
1–5 person team can build this	✅	Standard stack + LLM APIs + a curated rules pipeline. Beachhead-scoped.
Launchable with <$50K / ₹40L	✅	Software + data-curation labour; no capex.

All five pass.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	17/20	Hair-on-fire for the exporter who just got blacklisted; six/seven-figure single losses; felt per shipment. Slightly under 18 because not every exporter feels it every month.
Demand evidence	15	12/15	Hard RASFF numbers, documented losses, crowded doc-software market proving WTP. Gap: no direct evidence exporters will pay specifically for a screening layer vs assuming their consultant covers it.
Build feasibility	15	11/15	App is easy; the live multi-jurisdiction rules layer is the real work and must stay current.
Distribution clarity	15	11/15	Public exporter directory + commodity clusters + CHA channel + rejection-moment outreach. Named lists, but conversion of a “preventive” tool is uncertain.
Revenue mechanics	15	11/15	Rupee-micro pricing fits wallets; $1M path needs ~1,000 payers (7–12% of pool) — reasonable, not trivial.
Time to first revenue	10	7/10	A scoped one-lane MVP + free teardown funnel can land pilots in 6–10 weeks.
Defensibility	10	6/10	Moat = the accumulating, validated, always-current rules layer + audit-trail lock-in + trust after a rejection-free record. Copyable early, compounding later.
Total	100	75/100

13. Qualitative modifiers

Founder-fit tags

domain-expertise-required · technical-heavy

Needs someone who genuinely understands agri-export compliance (or a co-founder/advisor who does) plus the engineering to build the ingestion + reasoning pipeline. A pure generalist will get the rules subtly wrong — and “subtly wrong” here means a customer gets blacklisted, which is fatal to trust.

Key assumptions to validate (3–5)

1. Assumption: Small exporters will pay ₹3K–6K/mo for a preventive screen rather than rely on their existing consultant/lab. How to test: 30 exporter interviews + 10 paid pilots from the free-teardown funnel; track conversion and stated WTP.
2. Assumption: The destination rules layer can be kept current and accurate enough that a “GO” rarely precedes a real rejection. How to test: back-test the engine against 50 historical RASFF-rejected India consignments — does it flag them? Target ≥80% catch on documentary/MRL-mismatch failures.
3. Assumption: Exporters will upload a real CoA / lab report (trust + habit). How to test: in pilots, measure how many complete a full screen vs bounce at the upload step.
4. Assumption: The rejection-moment + directory outreach actually converts. How to test: run 200 personalised teardowns; measure reply and trial-start rate (kill if <5% reply).

Risk flags

1. Liability / false-assurance risk: if the tool says “GO” and the shipment is rejected, trust (and possibly liability) evaporates. Must be framed as risk-flagging with explicit boundaries — it screens documents and destination rules, it does not certify the chemistry. This is the single biggest product-design risk.
2. Data-currency / platform risk: rules change via dozens of jurisdictions’ irregular publications; a stale rules layer is worse than none. Ongoing curation cost is the real operating burden.
3. Scope creep into the lab’s job: the temptation to “also predict residues” leads into chemistry the team can’t own. Discipline required to stay in the documentary/rules-mismatch lane.
4. Channel dependency: CHA/forwarder partnerships are high-leverage but slow; over-reliance on them could stall direct growth.

14. Structured verdict

Score:                  75/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder + agri-export-compliance domain advisor (or ex-Spices Board / CHA insider)
Time to revenue:        6–10 weeks to first paid pilots (one-lane MVP)
Capital to launch:      ₹8–15 lakh ($10–18K) — mostly data-curation labour + LLM API
Top 3 assumptions to validate first:
  1. WTP for a preventive screen — 30 interviews + 10 paid pilots, kill if <3 convert
  2. Rules-engine catch rate — back-test vs 50 historical RASFF rejections, target ≥80%
  3. Teardown-funnel conversion — 200 personalised screens, kill if <5% reply
Kill criteria:
  - Abandon if <3 of 10 pilot exporters convert to paid after a free teardown
  - Abandon if the rules engine catches <60% of historical documentary/MRL-mismatch rejections in back-test
  - Abandon if a well-funded doc-software incumbent (Pazago et al.) ships an equivalent destination-rules screen before the v1 lands

15. Next step — 1-week validation sprint

• Day 1–2: Pull the EU MRL + reinforced-control rules for the top 5 India-flagged commodities (cumin, chilli, turmeric, black pepper, nutmeg). Build a throwaway screen that takes a commodity + a CoA’s residue numbers and returns pass/fail against the destination MRL. No UI — a notebook is fine.
• Day 3–4: Back-test it against 15–20 historical RASFF-rejected India consignments (the notifications are public). Does it flag the ones that should have failed? This is the falsifiable core: if it can’t catch known rejections, the idea is dead.
• Day 5: Take the working screen to 8–10 exporters from the APEDA directory (Unjha cumin cluster is the fastest concentration). Run their real last shipment through it live and ask: “would you pay ₹4,000/mo for this?”
• Decision rule: GO only if (a) the engine catches ≥80% of the back-tested rejections AND (b) ≥3 of 10 exporters say yes to paid at ₹4K/mo on the spot. Anything less → rework the wedge or pass.