LANDED pitch deck

Know Before You Trade.

TrendFlux Digital is the technology company. LANDED is its flagship product: AI-powered import intelligence that turns one supplier document into landed cost, HS classification, compliance and risk — before the money is committed.

• Category: AI-powered import intelligence · Outcome: smarter importing · Long-term: an AI Trade OS.
• Market focus: Bangladesh import economy, beachhead SME importers in Dhaka and Chattogram.
• Stage: pre-seed, pre-revenue, working product live on a custom domain.
• Startup World Cup Bangladesh 2026 · Founder: Zahid Hasan Emon.

Importers commit money before they know the real number

The decision to place an order happens before landed cost, HS classification and compliance exposure are actually known.

• Answers are assembled from Excel, WhatsApp, C&F agents, indenting houses and personal memory.
• Classification depends on one person's experience, so duty exposure is discovered late.
• Compliance requirements often surface at the port, not at the negotiating table.
• The true landed cost is confirmed after the goods are already moving — margin is eroded silently.

Source / note: Problem framing is founder-observed in the Bangladeshi import market. Structured importer interviews are an open validation item — see Evidence.

One pre-commitment trade analysis from one document

Upload a supplier invoice, quotation or proforma — or paste the trade details — and LANDED returns a single reviewable analysis.

• Structured line items, quantities, incoterms and values extracted from the document.
• Ranked HS code candidates with the reasoning behind each.
• Duty, VAT and fee estimates rolled into a landed cost and per-unit figure.
• Compliance checklist, trade risk signals, and a proceed / renegotiate / hold recommendation.
• Every figure labelled with its provenance: from document, AI estimate, or not in input.

A large, document-heavy import economy — and reasoning is finally cheap enough

Bangladesh imported roughly USD 67 billion of merchandise in FY2024-25, and nearly every consignment still starts as a proforma invoice a human reads by hand.

• China supplied USD 18.20bn (27.0%) and India USD 9.56bn (14.2%) of import payments in FY2024-25 (Bangladesh Bank).
• Existing software serves post-decision execution — freight, customs filing, accounting — not the pre-commitment decision.
• Multimodal models can now read a scanned proforma reliably enough for decision support, at a per-analysis cost an SME importer can afford.
• Importers are already digital on the communication layer (WhatsApp, email PDFs) but analogue on the decision layer.

Source / note: Source: Bangladesh Bank country-wise import payments, FY2024-25; the ~USD 67bn total is derived from the published China share (18,195.8m = 27.0%). The 'why now' technology argument is framing, not a sourced statistic.

Live working product, not a mockup

The judge can drive it in 90 seconds at /analyze on the live domain.

• Browser-side extraction for PDF, Word, image, TXT and CSV — documents are not stored on our servers.
• Server-side AI analysis with schema-validated output and recomputed totals.
• Verification checklist, CSV / JSON export and branded print-to-PDF report.
• Device-local analysis history. No account required to try it.

Pre-traction. What exists is product evidence.

• Current state: working product deployed on a custom domain with a verified end-to-end analysis flow.
• Current state: document ingestion, provenance labelling, verification checklist and export are implemented and demonstrable.
• Current state: zero users, zero pilots, zero letters of intent, zero revenue, no usage analytics, no accuracy benchmark — stated plainly.
• Future validation: AI output compared against a licensed broker's numbers on real invoices, then 3–5 structured pilots.

Source / note: We would rather show an honest zero than an invented number. Open evidence items are tracked on the Evidence page.

Beachhead SME importers — sizing method shown, figures labelled

Beachhead: Dhaka/Chattogram SME commercial importers who import monthly or more, receive supplier documents as PDF or image, and have no in-house trade compliance team.

• TAM: Bangladesh commercial importers who would pay for trade-decision software.
• SAM: regularly-importing SMEs reachable digitally in Dhaka and Chattogram.
• SOM: what a small pre-seed team can realistically serve in year one.
• Bottom-up method: importers × paid conversion × monthly price × 12 — inputs labelled Sourced fact / Derived / Assumption.
• Importer-count inputs are not yet verified against an NBR/BIDA registry, so no headline TAM figure is claimed.
• Expansion rings: trading companies, indenting houses, buying houses, manufacturer procurement teams.

Source / note: See Market Research for the full calculation, inputs, sources and confidence labels.

We sit before the decision, not after it

• Freight and customs platforms act after the buy decision; accounting tools act after the goods land.
• Global compliance databases sell data, not a decision; brokers sell judgement, not software.
• LANDED combines cost, classification, compliance and risk in one pre-commitment view built for emerging-market document reality (scanned, informal, incomplete).
• Intended compounding advantage: a Bangladesh-specific classification and landed-cost dataset built from real analyses — this does not exist yet.

Source / note: Moat is stated as intent. Today the defensibility is speed and market focus, not data.

Subscription per seat, plus usage above allowance

• Buyer: the importer's owner or procurement lead — the person who signs off the order.
• Paid for: confidence at the moment of commitment, and a shareable analysis artifact.
• Planned tiers: free trial analyses, a monthly SME plan, and a higher tier for indenting/team workflows.
• No billing is live and there are no paying customers. Pricing is intent, not traction.

Source / note: All pricing is planned. Unit-economics inputs are assumptions and are labelled as such.

Aggregator-led conversion, on a self-serve engine

• Reach importers through the intermediaries that already aggregate them: indenting houses, C&F agents, trade associations.
• Free first analysis on a real invoice as the entire top of funnel; founder-led, in-person, Bangla sales for the first cohort.
• 90-day plan targets validated pilots, not vanity signups.
• Scales without per-customer setup: a new importer runs a first analysis with no onboarding, integration or training.
• Marginal cost per analysis is model inference; the same document-to-decision engine extends to new corridors by changing reference data, not the core flow.

Source / note: Scalability is architectural today — not yet proven under real user load, because no usage data exists.

A balanced founding team

Technology + Finance + Execution — the three functions an early-stage AI trade-intelligence product needs.

Technology + Finance + Execution = A balanced founding team

Source / note: Roles are functional, not unverified titles: no CFO/CTO/Co-Founder claims. The founder's broader portfolio track record is personal history, not LANDED customer traction.

Funding Ask — BDT 50,00,000 (BDT 5 million ≈ US$41k)

We are raising a pre-seed product and market-validation round to move LANDED from a working product to paid, measured pilots. The USD figure is an approximate conversion and will shift with the exchange rate.

• Funding Ask: BDT 50,00,000 (BDT 5 million, approximately US$41k) pre-seed — a product and market-validation round, not a valuation claim.
• Use of funds: product/data infrastructure, authoritative trade and tariff data integrations, customer pilots and GTM, and team/operations.
• 12-month milestones: launch paid pilots, establish measurable usage and accuracy benchmarks, integrate authoritative tariff/HS data, validate repeatable customer acquisition.
• Non-capital ask: pilot introductions to importers and indenting houses, and a customs-broker validation partner.
• Vision: from AI import intelligence to digital AI indenting, and long term an AI Trade OS for cross-border trade intelligence and execution.

Source / note: Milestone-based use of funds; exact percentage allocations are deliberately not stated. No investors, revenue, users, pilots or LOIs are claimed.