Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

XIMVERSE

India's only self-serve CBEC-format export documentation engine. Fills Commercial Invoice + Packing List from your existing PDFs in under 2 minutes — no CHA, no re-typing. Built for the 63M MSMEs the ₹15,000-per-shipment documentation industry doesn't want to automate.

Primary OCR: OCR.space. Fallback: Tesseract.js + pdf.js (WebAssembly) — runs in-browser when quota runs out, no API key needed.

BITS Pilani Startup Hub Pilot: Deendayal Port Users Status License: MIT CI


🚀 Why XIMVERSE

India has 63 million MSMEs, yet fewer than 150,000 are active exporters. The #1 reason the rest don't ship abroad isn't product quality or demand — it's documentation.

Every shipment requires 10+ cross-border documents — Commercial Invoice, Packing List, Bill of Lading, Certificate of Origin, Shipping Bill, and more. Each document asks for the same exporter info: IEC, GSTIN, PAN, bank details. Small businesses waste 3–6 hours per shipment retyping the same data, paying CHAs ₹5,000–15,000 per consignment, and risking customs delays that cost lakhs from a single typo in an HS code.

XIMVERSE makes export documentation self-serve. An exporter uploads their company profile and source invoice once. Our OCR pipeline extracts every field. Structured JSON auto-fills standard CBEC-format templates. The exporter downloads customs-ready PDFs in under 2 minutes — no CHA, no re-typing, no errors.


🎯 The Real-World Problem We Solve

Who hurts today:

  • First-time exporters in Tier-2/Tier-3 cities who don't know where to start
  • Small textile, agri, and handicraft manufacturers in Surat, Tiruppur, Jaipur, Moradabad, etc.
  • Women-led and rural MSMEs entering global trade
  • Repeat exporters tired of paying agents for work that's 95% copy-paste

What they told us (from 183 early users):

  • 73% said documentation is their biggest bottleneck — bigger than finding buyers
  • Average exporter spends ₹3,000–8,000 per shipment on CHA fees just for docs
  • 1 in 4 shipments faces delays due to documentation errors
  • 89% said they'd use a self-serve tool if it mapped to standard templates

📊 Traction & Validation

  • 183 active users across 14 Indian states
  • Pilot deployment underway at Deendayal Port Authority (Kandla) for real-shipment testing alongside this build
  • Empanelment in progress with the Ministry of Ports, Shipping & Waterways as a service provider
  • Incubated at BITS Pilani Startup Hub
  • Backed by BITS Pilani Startup Hub (grant-supported)
  • Active user feedback loop driving feature roadmap

✨ What's Built (MVP — this repo)

The current codebase is the core automation engine powering the pilot:

  • ✅ Upload Commercial Invoice PDF → OCR.space extracts shipment data
  • ✅ Upload Company Profile PDF → populates exporter profile
  • ✅ One-click "Create Consignment" → both PDFs auto-generated
  • ✅ Download customs-ready Commercial Invoice + Packing List as a ZIP
  • ✅ Consignment history with re-download
  • ✅ Trade insights panel (currency, port alerts, export trends)

🆚 Why Not [Docparser / Nanonets / ImpEx / generic doc tools]?

The competition XIMVERSE
Generic document automation — works for any country, any format India-specific — CBEC format with IEC, GSTIN, Shipping Bill, HS code fields pre-wired
LLM-based extraction — expensive per request, hallucinates field values Deterministic regex anchored to fixed labels — free, instant, never hallucinates a bank account number
API-dependent OCR — platform goes down, quota runs out, your product breaks WebAssembly fallback (Tesseract.js + pdf.js in-browser) — zero dependency on any external API when it matters
Enterprise SaaS priced for freight forwarders MSME-first — free tier, demo user bypass, no signup friction. Built for a first-time exporter in Surat, not a CHA's back office
English-only, global market focus Tier-2/3 India focus — Deendayal Port pilot, DPDP Act compliance, FIEO/APEDA partnership roadmap

🏰 The Moat — Why XIMVERSE Wins This

India's ₹28 lakh crore export industry still runs on Excel.

No hyperbole. Commercial Invoices are typed by hand into Word templates. Packing Lists are copy-pasted from the last shipment. HS codes are looked up in PDFs. Errors cause customs holds that cost ₹2–5 lakh per incident. There is no AI-powered, self-serve export documentation tool built for Indian MSMEs. Zero. The incumbents — Customs House Agents, freight forwarders, trade consultants — profit from this complexity and have no incentive to automate it away. We do.

The wedge: India's agri-commodity SME exporters

XIMVERSE's entry point is small and mid-sized exporters in India's agri-commodity sector — rice, spices, pulses — who face three compounding problems:

  1. Manual documentation — every shipment means retyping the same exporter IEC, GSTIN, and bank details into a fresh Word file
  2. Fragmented workflows — Commercial Invoice, Packing List, B/L data, and LC terms all live in separate files with no single source of truth
  3. No access to ERPs — SAP and Oracle cost ₹10–50 lakh to implement; Tally doesn't do export-specific CBEC formats

Unlike generic document tools, XIMVERSE is purpose-built for export-specific workflows — Invoice, Packing List, B/L data, LC terms — with pre-mapped templates aligned to DGFT guidelines and real trade practices. This niche is underserved, highly repetitive, and sticky: once an exporter's IEC, GSTIN, bank details, and product HS codes are in the system, they never leave.

Why the moat is real and defensible

Layer Why it's hard to copy
Undigitized segment, zero competition The only "competition" is Excel, WhatsApp-forwarded Word files, and ₹15,000-per-consignment CHAs. No funded startup is targeting this.
Purpose-built for export workflows Pre-mapped CBEC templates (Invoice, Packing List, B/L, LC terms) aligned to DGFT — not adapted from a generic doc tool
India-specific compliance formats IEC codes, GSTIN validation, Shipping Bill (ICEGATE), DGFT export license fields — a foreign SaaS can't clone this without years of domain work
First-mover relationships Deendayal Port Authority pilot, BITS Pilani Startup Hub, FIEO/APEDA pipeline — institutional trust takes years to build and can't be bought
AI roadmap nobody has started The regex engine is v1. The AI layer — automatic HS code classification, compliance risk flagging, hallucination-free extraction from any document format — is v2 and nobody has built it for this segment
Government portal integration path ICEGATE (customs filing), DGFT (export licenses), FSSAI, MoC — becoming the trusted connector between SMEs and government portals is a decade-long position

Expansion path: wedge → ecosystem

v1 (now)      Agri-commodity SME exporters (rice, spices, pulses)
              2 documents, CBEC format, deterministic extraction
v2 (Q3 2026)  All MSME commodities, 10-document pack, AI-assisted fill
              Importers + CHAs added as user types
v3 (2027)     API layer for freight forwarders and CHA back-offices (B2B SaaS)
              ICEGATE real-time filing, compliance risk scoring
Platform      1-click complete export pack for every participant:
              exporters → importers → CHAs → freight forwarders →
              inspection agencies → government portals

The wedge is agri-commodity SME exporters. The platform is the entire Indian cross-border trade ecosystem — a market with no dominant software player, 63 million businesses waiting to be unlocked, and export documentation that has not changed in 30 years.


🏗️ Tech Stack

Novel Engineering Decisions

Decision Why
WebAssembly OCR fallback When OCR.space quota is exhausted, we fall back to a client-side WASM pipeline: pdfjs-dist renders PDF pages to <canvas>, then Tesseract.js (WebAssembly) extracts text. Zero server round-trip, zero API key, zero quota.
Regex over LLM Trade docs have fixed, labeled fields. Regex is free, instant, and deterministic — no hallucination risk on IEC/GSTIN/bank fields. LLM was prototyped and dropped.
Browser-direct OCR OCR.space is called from the browser, not the server. We discovered Vercel's 10s serverless timeout was killing extractions on larger PDFs — moving it client-side eliminated the failure entirely.
Supabase over custom backend Auth + DB + storage in one, letting us focus on the pipeline. We evaluated Firebase but chose Supabase for SQL and row-level security on user documents.

Tech Stack

Layer Choice
Framework Next.js 14 (App Router) + TypeScript
Styling Tailwind CSS
Auth Supabase Auth (+ demo user bypass for evaluators)
Database Supabase PostgreSQL
Storage Supabase Storage
Primary OCR OCR.space API (browser-direct)
Fallback OCR Tesseract.js + WebAssembly (client-side, no API key)
PDF Rendering pdfjs-dist (WebAssembly, for Tesseract fallback pipeline)
JSON Parsing Plain JavaScript (regex-based, deterministic)
PDF Generation @react-pdf/renderer
Bundling jszip
Deployment Vercel

🧭 Core Workflow

1. Upload Company Profile PDF (once)
   └─ OCR.space extracts company data → saved to profiles table

2. Upload Commercial Invoice PDF (per consignment)
   └─ OCR.space extracts shipment data → parsed to structured JSON

3. Click "Create Consignment"
   └─ Profile + invoice JSON merged → Commercial Invoice + Packing List PDFs generated in-browser

4. Download ZIP
   └─ Both CBEC-format PDFs bundled → customs-ready in under 2 minutes

The parsing engine uses deterministic regex anchored to fixed field labels (IEC, GSTIN, Invoice No, B/L No, etc.) — no LLM, no API cost, no hallucinations. Fallback data ensures the demo always runs even if OCR misses a field.


🚀 Quick Start

Option A — Use the live demo (no setup)

https://ximverse.vercel.app

Click "Continue as Demo User" — no sign-up, no credentials. A sample consignment is pre-loaded.

Option B — Run locally

git clone https://github.com/ashutoshrai7205/XimVerse.git
cd XimVerse
npm install

cp .env.example .env.local
# Fill in:
#   NEXT_PUBLIC_SUPABASE_URL=
#   NEXT_PUBLIC_SUPABASE_ANON_KEY=
#   OCR_SPACE_API_KEY=

npm run dev
# → http://localhost:3000

You'll need:

  • A free Supabase project (schema in supabase/)
  • A free OCR.space API key (500 requests/month)

🔒 Privacy & Data Safety

We take data privacy seriously — especially because our users upload sensitive business documents containing GSTIN, IEC codes, and bank details.

Practice Detail
Storage Uploaded PDFs stored in private Supabase Storage buckets; access-controlled per user_id — no cross-user leakage
Retention Raw OCR text is not persisted; only the structured JSON extracted from it is stored
No data brokering User data is never sold, shared, or used for advertising
Deletion Users can delete their profile and all consignment data at any time from the dashboard
Encryption All data encrypted in transit (TLS 1.3 via Vercel + Supabase) and at rest (AES-256 via Supabase)
Compliance Built in alignment with India's Digital Personal Data Protection Act, 2023 (DPDP Act)
Third parties The only third-party processor is OCR.space; document text is sent for OCR only — not retained by them under their free-tier terms

💰 Pricing & Sustainability

XIMVERSE is designed to be financially self-sustaining within 6 months of the pilot.

Plan Price Consignments/month Target
Free ₹0 5 First-time exporters trying the platform
Starter ₹499/month 50 Small exporters, 1–4 shipments/week
Business ₹1,999/month Unlimited Growing SMEs, freight forwarders
Enterprise Custom Unlimited + API CHAs, trade associations, port authorities

Revenue beyond subscriptions:

  • White-label licensing to Customs House Agents (CHAs)
  • API access for ERP integrations (Tally, Busy, Zoho Books)
  • Partnership with FIEO/APEDA for member-subsidised access
  • Port authority SaaS contracts (Deendayal pilot → template for replication)

🗺️ Roadmap

Q2 2026 — Core Pack Expansion

  • GST Tax Invoice generation
  • Certificate of Origin (Form A, SAPTA)
  • Shipping Bill pre-fill (ICEGATE format)

Q3 2026 — Integrations

  • REST API for ERP connectors (Tally, Zoho Books)
  • Direct ICEGATE e-filing integration
  • Bulk consignment processing (CSV upload)

Q4 2026 — Scale & Mobile

  • Mobile app (React Native, iOS + Android)
  • Multi-user team accounts with role-based access
  • Audit trail and e-signature support (DSC)

2027 — Platform Play

  • Full 10-document export pack
  • CHA partner network (reseller programme)
  • AI-assisted HS code classification
  • Real-time customs duty calculator

🤝 Contributing

This project is in active development as part of a pilot with Deendayal Port Authority. Contributions are welcome — especially from:

  • Exporters who want to report edge cases in OCR parsing
  • Developers familiar with CBEC document formats
  • Anyone who has shipped goods and knows how painful this is

To contribute:

  1. Fork the repo
  2. Create a feature branch: git checkout -b feat/your-feature
  3. Run tests: npm test
  4. Submit a pull request with a clear description

Please open an issue first for significant changes.


📄 License

MIT — free to use, modify, and distribute. Attribution appreciated.


Built at BITS Pilani Startup Hub. Solving real problems for real exporters.

Impact

What's visible today is a demo. What we're actually building is infrastructure.

The core insight Export documents are not paperwork — they are structured information that ports, customs, and logistics systems consume to automate operations. Every delay at an Indian port today is, at its root, a document-handling delay. Fix the document layer and you unlock everything downstream. Operational impact — turnaround time

Port turnaround time drops when documents arrive clean, validated, and machine-readable. Customs clearance accelerates because data moves directly between systems instead of being re-keyed. Fewer manual handoffs means fewer errors, fewer demurrage charges, and faster vessel cycles.

Ease of doing business — unlocking the long tail of exporters Today, export/import is effectively gated by documentation complexity. A handicraft artisan in Jaipur, a weaver in Varanasi, or a small food processor in Kerala cannot realistically navigate the 30+ document types, multi-agency approvals, and compliance formats required to ship abroad. Our platform collapses that complexity into a guided, automated workflow — making export accessible to small producers who have always been export-capable but never export-enabled. Macro-economic flywheel This is why it matters at a national scale:

More exporters → India's export base diversifies beyond a handful of sectors and large players. Diversified exports → reduced vulnerability to sector-specific shocks; stronger forex position. Rising export volume → manufacturing sector expands to meet demand. Manufacturing growth → job creation across tier-2 and tier-3 cities. Jobs → disposable income in new hands. Income → domestic demand for better, more innovative products. Demand → pressure on manufacturers to innovate and upgrade. Loop closes — and the economy compounds.

We are solving the bottleneck at step 1 so the rest of the flywheel can actually spin. Strategic alignment and credibility

Addresses 9 of the 20 "must-have" solutions identified under the India Maritime priority list. Accepted by DGFT (Directorate General of Foreign Trade), Ministry of Commerce — meaning the approach is validated at the policy level, not just the product level.

Architectural positioning — the "UPI moment" for trade The analogy we keep coming back to:

PhonePe and Google Pay did not build a payments network. They built the experience layer on top of NPCI's UPI rails.

That is exactly what we are doing for maritime trade.

NLP Marine is the underlying rails — the protocol layer for port and logistics data. We are the orchestration layer that sits on top — connecting service providers, exporters, customs, and ports into a single automated flow. Just as UPI unlocked a decade of fintech innovation by abstracting away banking complexity, a clean orchestration layer over MLP Marine can unlock a decade of trade-tech innovation in India.

In one line We are not digitising documents. We are building the orchestration layer that turns India's maritime trade rails into a platform — so that the next million exporters can plug in, and the economy can compound on top.

About

Web app for export traders to automate documentation. Upload source PDFs → OCR extracts shipment data → templates auto-fill → Commercial Invoice & Packing List generated and bundled for download. Built with Next.js 14, Supabase, and OCR.space.

Topics

Resources

Contributing

Security policy

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages