MET Holding
AI-Native Document & Verification Operations
Dubai-based trade finance and corporate services firm handling high-value cross-border instruments, SBLC packages, bank guarantees, and partner-bank submissions.
processed in first quarter
days verification cycle
bank acceptance rate
less manual screening
From document chasing to a self-running verification layer.
AI-native verification operation
15 days
verification cycle
$250M+
first-quarter volume
85%
bank acceptance
3 days
audit prep
MET Holding's work is document-heavy by nature. Each transaction depends on identity records, corporate files, financial statements, bank references, ownership documents, and instrument-specific paperwork.
Before REPCONN, verification meant long manual review cycles, repeated document chasing, and bank submissions that could fail because of small but expensive gaps.
REPCONN rebuilt the workflow as an AI-native document and verification operation. The system reads, parses, screens, validates, and organizes transaction evidence while people handle only judgment calls and escalations.
High-value trade finance was bottlenecked by paperwork.
Every transaction was a document project
SBLCs, bank guarantees, letters of credit, corporate documents, passports, financial statements, property files, and bank references had to be collected, checked, compared, and formatted before submission.
Verification was slow and expensive
A single transaction could require 30+ days of manual review across jurisdictions, document versions, stakeholder emails, and partner-bank requirements.
Bank submissions failed on avoidable gaps
Each partner bank had different expectations. Missing fields, inconsistent ownership data, outdated documents, and formatting issues created rejection-resubmission cycles.
Audit preparation took weeks
Files were spread across email, shared drives, and manual trackers. Preparing a clean audit package could take three weeks because every document action had to be reconstructed.
Our Solution
One operating layer for documents, risk, bank readiness, and audit.
REPCONN connected document AI, computer vision, KYC/AML checks, sanctions screening, bank-specific validation, secure storage, and audit trails into one managed workflow.
System 01
Document Intelligence Layer
REPCONN built a document processing layer that reads passports, corporate certificates, financial statements, and trade finance packages across five languages.
What changed
Computer vision for scanned documents
AI extraction for names, entities, dates, values, and identifiers
Structured transaction files from unstructured uploads
Document completeness checks before submission
Document engine
Passports
Passport scans and identity documents are read, extracted, and compared against client records and screening requirements.
System 02
KYC, AML, and Sanctions Screening
The verification workflow screens people, entities, beneficial owners, and transaction patterns before bank submission, with escalation paths for judgment-heavy cases.
What changed
KYC and beneficial ownership checks
PEP, adverse-media, and sanctions screening
Risk flags for enhanced due diligence
Screening aligned to FATF and US sanctions standards
Verification loop
Extract
Computer vision and document AI extract names, entities, ownership details, dates, values, and document identifiers.
System 03
Bank-Ready Submission Workflow
REPCONN standardized transaction packages so MET Holding could prepare cleaner first submissions and reduce costly bank rejection cycles.
What changed
Bank-specific documentation checklists
Missing-field and inconsistency alerts
Version control for every transaction file
Submission packages prepared from one source of truth
System 04
Secure Audit Repository
All transaction materials, review actions, approvals, and document changes are kept in a secure repository designed for fast internal review, partner-bank questions, and audit preparation.
What changed
AES-256 encryption and role-based access
Full audit trails for file access and approvals
Transaction history retrievable in minutes
Audit prep reduced from three weeks to three days
Faster verification, cleaner bank packages, and more volume through the same team.
Processed in first quarter
The system supported more than $250M in high-value trade finance transaction volume in the first quarter after deployment.
Verification cycle
Average verification time moved from roughly 30 days to 15 days by automating document extraction, completeness checks, and screening work.
Bank acceptance rate
First-submission acceptance improved from 70% to 85% as document packages became cleaner, more complete, and more consistent.
Less manual screening time
Routine sanctions, identity, and document screening shifted from manual review to an AI-native verification workflow.
Audit prep time
Audit preparation dropped from three weeks to three days because transaction files, approvals, and evidence trails became instantly retrievable.
Languages supported
Document AI and computer vision process trade finance, identity, and financial documents across five languages.
Compliance built into the document workflow.
The operating layer was built with AES-256 encryption, role-based access, full audit trails, KYC/AML screening, PEP and adverse-media checks, and sanctions screening aligned to FATF and US sanctions standards.
Verification loop
Extract
Computer vision and document AI extract names, entities, ownership details, dates, values, and document identifiers.
Real SBLC work happens through banks, with MET coordinating the document operation.
A standby letter of credit is a bank-issued undertaking used to support payment or performance obligations. MET Holding was not the issuing bank and did not replace the bank-to-bank process. Its role was to coordinate the client-side transaction file: collecting documents, preparing bank-ready packages, managing verification, and supporting the compliance workflow around the instrument.
SBLC work remains bank-to-bank: the applicant requests the instrument, the issuing bank sends the authenticated SWIFT message, the receiving or advising bank receives it, and the beneficiary relies on the bank undertaking. MET Holding's workflow required strict document verification, bank-ready files, and direct coordination around real banking rails.
REPCONN supported workflows involving major institutions, including Deutsche Bank as issuing bank and First Abu Dhabi Bank as receiving bank, with controls designed for SWIFT-compatible documentation and review.
Bank-to-bank flow
Applicant
Buyer or borrower requests SBLC
MET Holding
Coordinates client-side documents, verification, and bank-ready transaction file
Issuing Bank
Issues the undertaking
SWIFT MT760
Authenticated SBLC message
Receiving Bank
Receives or advises the instrument
Beneficiary
Seller or lender relies on the bank undertaking
MT760
Issues the SBLC and creates the bank undertaking.
MT799
Authenticated free-format readiness or information message. It is not an SBLC.
MT767
Amends an existing guarantee or SBLC, such as amount, expiry, or terms.
The verification operation runs continuously. People only step in for judgment calls, escalations, and relationship decisions.
Confidentiality Notice
Specific partner-bank requirements, client identities, transaction files, risk rules, and internal system configurations are not published. Metrics represent rounded public-facing outcomes from the engagement.
Get Started
Want a document-heavy operation to run like this?
We assess your current document workflow, identify where AI-native systems will have the highest impact, and build the operating layer around your risk, audit, and revenue requirements.