SB StartupBasket
All ideas
75 /100 GO Medium complexity

PresentPerfect — LC pre-check for India's small exporters

Reads your export document set against the LC and UCP 600, flags every discrepancy before the bank does.

TradeTech SaaS India WhatsApp-firstSMBAI-agentCompliance-drivenMultilingual
— views
Evaluation Scores
75/100

GO

Overall Score

17
Problem
12
Demand
11
Build
11
Distrib.
12
Revenue
7
Time
5
Defense

PresentPerfect

1. One-liner

Reads your export document set against the LC and UCP 600, flags every discrepancy before the bank does.

2. Trend signal — why now?

Three things converged in the last twelve months:

  1. The pain is measured and brutal. 60–75% of LC document presentations are rejected on first submission for UCP 600 non-compliance, and banks charge a discrepancy fee for each error. A single Bill-of-Lading vs LC mismatch can freeze ₹15,00,000 of cargo at the destination port. When docs are discrepant, the buyer is asked to waive — and routinely uses that waiver as leverage to renegotiate the price down. This isn’t a paperwork annoyance; it’s money walking out the door.

  2. The tech got cheap enough to read the whole document set. Reading an invoice, Bill of Lading, packing list, insurance certificate and the LC text (Field 47A conditions and all), then cross-checking every field against UCP 600 rules and the LC terms, is exactly the kind of multi-document reasoning that became reliable and cheap in 2025–26. Two exporter-side tools (SmartLC in the UK, Loamist for enterprise) launched on precisely this unlock.

  3. The money and the market are both moving. AI usage in live trade-finance transactions rose from 32% (2024) to 45% (2025). India’s MSMEs generate ~48.6% of merchandise exports (₹9.52 lakh crore in H1 FY26 alone) across 7.8 crore registered Udyam enterprises — a huge base of small exporters who use LCs but can’t afford a Loamist enterprise contract or a full-time documentation manager.

Provenance:

3. The opportunity

The gap is a wallet-and-geography arbitrage, not a new market.

Today a small Indian exporter has three options when a shipment is on an LC: (a) eyeball the documents themselves and pray, (b) pay their bank’s trade-desk to check — which happens after presentation, when it’s too late and the fee is already ticking, or (c) hire an export-documentation consultant / CHA at consultant rates for a pre-check. Most pick (a) and eat the ~65% rejection rate.

The AI checkers that solve this properly are aimed elsewhere. Loamist is explicitly enterprise — built for chemical/commodity houses doing 100+ LCs a month on $150K–$750K values, sold via demos and multi-week deployments. SmartLC is UK-based and generic — international, English-doc-inbox, no India localization, no rupee pricing, no WhatsApp. Neither will chase a Tirupur textile exporter doing 8 LCs a month, because the enterprise sales motion doesn’t pay for that customer.

That customer is exactly who’s bleeding. A focused, India-first, rupee-priced, WhatsApp-native pre-check that catches the discrepancy before the documents ever reach the negotiating bank is a 10× better answer for them than “check it after the bank refuses.”

4. Target market

  • Primary customer: Owner or export-ops person at an Indian MSME exporter — textiles/garments (Tirupur, Surat), engineering goods (Ludhiana, Rajkot), spices/agri (Kochi, Guntur), handicrafts, pharma-intermediates — doing roughly 5–40 LC-backed shipments a month, ₹5–100 cr annual export turnover, 0–1 dedicated documentation staff.
  • Why they buy (in their words): “The bank rejected our set over a date mismatch on the BL and now the buyer wants a 4% discount to waive it.” “One discrepancy froze ₹15 lakh of cargo at the port.” They buy to stop payment from getting held hostage.
  • Rough TAM reasoning: India has lakhs of active merchandise exporters; even a conservative 150,000–250,000 SMEs use LCs regularly. Capture 3,000 paying at a ₹-tier and you’re comfortably past ₹10 cr ARR without touching the enterprise segment.
  • Why now for them: First-pass rejection rates are public and painful, buyers increasingly weaponize the waiver to squeeze price, and — critically — cheaper exporter-side AI tools now exist, so “there’s no software for this” is no longer true. The only reason they haven’t bought is that the existing tools aren’t built or priced for them.

5. Product sketch (MVP)

  • Upload the set + the LC (PDF/photo/scan via web or WhatsApp): invoice, BL/AWB, packing list, insurance cert, certificate of origin, plus the LC / MT700 text.
  • Discrepancy report in minutes, ranked by severity, each flag written in plain language: “BL shipped-on-board date 12 May is after LC latest-shipment date 10 May — this WILL be refused.”
  • Four-layer check: the LC’s explicit terms, UCP 600 implied rules, Field 47A additional conditions, and cross-document consistency (name spellings, amounts, quantities, dates, HS codes, Incoterms).
  • India-specific checks: shipping-bill value vs invoice value match (Indian customs reassessment trigger), e-BRC / realization consistency, common Indian-bank presentation quirks.
  • Fix-it guidance, not just flags: exactly what to amend and whether it needs a fresh document or an LC amendment request.
  • Presentation deadline tracker: counts down the LC’s presentation window so nothing lapses (late presentation is ~15% of all discrepancies).
  • WhatsApp delivery: send the docs in, get the report back in the same thread the ops person already lives in.

6. AI angle — what’s load-bearing

Remove the AI and there is no product. The core is reading five-to-eight unstructured, inconsistently-formatted documents (many scanned or photographed), extracting the ~80–150 relevant fields, then reasoning about them against a rule set (UCP 600) and a per-deal contract (the specific LC’s terms and Field 47A conditions) to find contradictions. That’s document understanding + multi-document cross-reference + rule application — not a form with validations. A hard-coded rules engine can’t handle “the LC says ‘Cotton Shirts’ but the invoice says ‘Men’s Woven Tops’ — is that a description discrepancy?” The judgment call is the product.

7. Localization angle

This is the whole thesis. India-first is the wedge, not a nice-to-have:

  • Pricing: a ₹2,000–6,000/mo tier works where SmartLC/Loamist’s dollar enterprise pricing can’t. The target customer will never sign a $1,500/mo demo-driven contract.
  • Distribution channel: WhatsApp-native intake and reporting — the actual surface Indian export-ops runs on.
  • Local doc set + rules: shipping-bill value matching, e-BRC realization, GSP/Certificate-of-Origin formats, and the specific presentation habits of Indian negotiating banks — none of which a UK-generic tool tunes for.
  • Language: report summaries in Hindi/Gujarati/Tamil for the owner even when the docs are English.

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

  • Pricing: ₹2,999/mo (up to 10 LC checks) → ₹5,999/mo (up to 30) → ₹9,999/mo (up to 75), plus ₹399/check pay-as-you-go for occasional exporters.
  • ACV: ₹48,000 ($575) at the mid tier.
  • Rough math to $1M ARR (~₹8.3 cr): ~1,750 customers at ₹48K ACV. Very reachable inside the SME exporter base.
  • Rough math to $5M ARR (~₹42 cr): ~7,500 mid-tier customers, or fewer with an upsell layer (see below). Still an order of magnitude below the LC-using SME population.
  • Expansion path: usage-based check volume growth; add adjacent modules — LC amendment-request drafting, e-BRC/realization tracking, buyer-name/sanctions screening, and a “documentation-as-a-service” done-for-you tier at ₹15K+/mo for exporters with zero ops staff.

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

  • Export Promotion Councils are pre-aggregated lists. AEPC (apparel), EEPC (engineering), Spices Board, GJEPC, Handloom EPC — each publishes/knows its member roster. Run a free “LC discrepancy audit” clinic with one council in one cluster (start with Tirupur/AEPC), check 50 real sets live, convert the ones you save from a rejection.
  • CHAs and freight forwarders as channel partners. They already sit between the exporter and the docs and get blamed when a set is rejected. Give them a partner tier — they push PresentPerfect to their exporter clients as “we now pre-check your LC docs,” you rev-share.
  • The pain-thread cold outreach. Indian trade forums (CAclubindia, TaxTMI, LinkedIn export groups) are full of “bank rejected our LC set” posts. DM the poster: “we’d have caught that BL date before you presented — send us your next set free.” High-intent, high-conversion.
  • Bank trade-desk relationship managers informally refer frustrated SME clients; a small referral incentive turns the party that causes the pain into a distributor.

10. Build complexity — justification

Medium. The document-reading and UCP-600 cross-check runs on off-the-shelf models and standard web + WhatsApp Business API plumbing — no custom infra. The real work is the domain layer: encoding UCP 600 + ISBP practice, the India-specific checks, and getting extraction reliable on photographed/scanned documents. That needs a trade-finance advisor and a corpus of real (redacted) LC sets to tune against, which is why it’s a Medium 3–4 month build, not a 6-week one.

11. Gating checklist

GatePass?Note
Legal in target market✅It’s a decision-support checker; the bank remains the authority. No license needed.
Ethical — no harm / dark patterns✅Helps exporters get legitimately paid; no manipulation.
Market exists (evidence above)✅60–75% first-pass rejection, ₹15L cargo freezes, two funded incumbents.
1–5 person team can build this✅1 technical + 1 domain advisor for v1.
Launchable with <$50K / ₹40L✅Off-the-shelf models + WhatsApp API; main cost is the advisor’s time.

All five pass.

12. Feasibility score

AxisWeightScoreNotes
Problem intensity2017/20Hair-on-fire: money frozen, buyer weaponizes waivers, ~65% first-pass fail. Felt on every LC shipment.
Demand evidence1512/15Public rejection stats, ₹15L freeze case, two live exporter-side incumbents. Docked because SMB-India willingness to pay for pre-check specifically is inferred, not yet proven.
Build feasibility1511/15Off-the-shelf models + WhatsApp, but scanned-doc extraction + UCP-600 domain encoding is real work; 3–4 months.
Distribution clarity1511/15EPC clinics, CHA channel, pain-thread outreach are concrete; council/CHA partnerships add coordination overhead.
Revenue mechanics1512/15Clear rupee tiers benchmarked below enterprise incumbents; path to $1M needs ~1,750 customers — reachable but not trivial.
Time to first revenue107/10Free-audit-to-paid can convert in 4–8 weeks once the checker is live; not instant because trust-building on real money matters.
Defensibility105/10Execution + India-tuned rule set + CHA/council relationships. SmartLC/Loamist could move downmarket; moat is focus and localization, not tech.
Total10075/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required — needs someone who can ship reliable multi-doc extraction AND a trade-finance advisor who’s lived UCP 600 rejections.

Key assumptions to validate (3–5)

  1. Assumption: Small Indian exporters will pay ₹3–6K/mo to pre-check LC docs (vs. eating the rejection or leaning on their CHA for free). How to test: Run the AEPC/Tirupur free-clinic, check 50 real sets, then ask the ones you saved to pre-commit at the tier price.
  2. Assumption: AI extraction is reliable enough on photographed/scanned Indian export docs to catch the discrepancies that actually get sets refused. How to test: Assemble 100 real (redacted) sets with known bank outcomes; measure catch rate vs. the bank’s actual refusal notice.
  3. Assumption: CHAs/forwarders will channel-partner rather than see this as a threat to their doc-prep fees. How to test: Pitch 10 CHAs the rev-share partner tier; count how many sign a pilot.
  4. Assumption: The India-specific checks (shipping-bill/invoice value match, e-BRC) are a real differentiator SMEs value, not just a feature. How to test: A/B the pitch with vs. without the India module against 30 exporters.

Risk flags

  1. Incumbent down-market move: SmartLC or Loamist could launch a rupee tier and WhatsApp intake. Mitigation is speed + council/CHA relationships they won’t build.
  2. Liability/trust: If the tool misses a discrepancy and a set is refused, the exporter blames the tool. Must be positioned as decision-support with clear “bank is final authority” framing, and catch-rate must be genuinely high before charging.
  3. Data cold-start: Needs real LC sets to tune the domain layer; without an advisor’s network, extraction quality lags and the free clinic underwhelms.

14. Structured verdict

Score:                  75/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder + trade-finance/UCP-600 domain advisor
Time to revenue:        6–10 weeks (free clinic → paid conversion)
Capital to launch:      ₹6–12 lakh ($7–14K)
Top 3 assumptions to validate first:
  1. SMB exporters pay ₹3–6K/mo for pre-check — AEPC/Tirupur free clinic, pre-commit the saved ones
  2. Extraction catch-rate on real scanned Indian sets — 100 sets with known bank outcomes
  3. CHAs partner not compete — pitch 10, count pilots
Kill criteria:
  - Abandon if catch-rate on the 100-set benchmark is below ~85% of what the bank's refusal notices flagged
  - Abandon if <10% of exporters saved in the free clinic convert to a paid tier within 30 days
  - Abandon if SmartLC/Loamist ship a rupee-priced WhatsApp India tier before your v1

15. Next step — 1-week validation sprint

  • Day 1–2: Get 20–30 real (redacted) LC document sets with known bank outcomes — from a CHA contact or an EPC member. This is the whole ballgame; without real sets you’re guessing.
  • Day 3–4: Run them through a hand-assembled checker prototype (off-the-shelf model + a UCP-600 checklist). Measure: of the discrepancies the banks actually flagged, how many did we catch — and how many false alarms did we throw?
  • Day 5: Decide go/no-go on a single falsifiable number: did we catch ≥85% of the real refusal-notice discrepancies with a tolerable false-positive rate? Below that, extraction/domain quality isn’t there yet and the free clinic will flop — no-go until it is.

The result is falsifiable: a catch-rate percentage against known bank outcomes, not “exporters seemed interested.”

Interested in a detailed proposal?

Get a deep-dive with market research, competitive analysis, and implementation roadmap.

Contact us

info@startupbasket.ai