AI-powered intercompany reconciliation that actually works.

Continuous matching across entities and jurisdictions. Deploy in weeks, not years.

Built by treasury practitioners. Currently onboarding design partners.

Trezh — Reconciliation Hub
Acme Global — Intercompany Matching
Live
Matched Today
142
↑ 94.7% auto-matched
Pending Review
8
3 FX variance
Entities Active
13
7 jurisdictions
Entity Pair
Corridor
Amount
Confidence
Status
Acme Brazil Acme Argentina
BRL → ARS
$2.4M
98.2%
Matched
Acme Mexico Acme Colombia
MXN → COP
$890K
96.5%
Matched
Acme Brazil Acme Korea
BRL → KRW
$1.7M
78.1%
FX Review
Acme HQ Acme Brazil
USD → BRL
$5.1M
99.1%
Matched
The Problem

Your spreadsheet is not a reconciliation engine.

Every month, your intercompany reconciliation process looks the same: pull SAP reports, cross-reference bank statements, chase down counterparty confirmations over email, and manually match transactions in a spreadsheet that grows more fragile every quarter.

Discrepancies surface at month-end close — days or weeks after the underlying transactions settled. By then, the cost of investigation has compounded, and the audit trail is a patchwork of email threads and Excel tabs.

Legacy TMS platforms promise to fix this, but they take 12–18 months to deploy, cost millions, and still process in batch. You don't need another system bolted onto your ERP. You need intercompany reconciliation that works the way your transactions actually flow — continuously, across entities, in real time.

The Reconciliation Hub

Continuous. Instant. Auditable.

Continuous Matching

Transactions match as they happen — not in monthly batch cycles. When Entity A in São Paulo records a payable and Entity B in Buenos Aires records the receivable, Trezh links them immediately with AI-powered confidence scoring.

Instant Discrepancy Detection

Amount mismatches, FX rate variances, and missing counterparty legs surface within minutes — not at month-end close. Your team investigates while context is fresh, not weeks later when nobody remembers the details.

Auditable Trail

Every transaction, match decision, state change, and correction is recorded in a tamper-proof shadow ledger. No more reconstructing what happened from email threads and spreadsheet versions. Your auditors will notice.

Integration

Trezh sits between your ERP and your banks. It doesn't replace either.

SAP / Oracle ERP
GL entries
Entity master data
IC agreements
TREZH
Shadow Ledger
Reconciliation Engine
Compliance Docs
Banking Partners
Bank statements
Transaction data
Settlement confirms

Your ERP remains your general ledger. Your banks remain your transaction processors. Trezh reads from both, assembles a unified intercompany transaction view, and writes reconciliation results back. The money continues to flow through your existing channels.

Day One

Scheduled extraction from SAP/Oracle + bank statement ingestion. File-based. No IT project required.

Over Time

Automated file delivery → full bidirectional API. Your integration deepens as your team's confidence grows.

Why Trezh

Built by people who've lived this problem.

Trezh was founded by operators who spent careers inside the treasury departments and technology organizations of the world's largest consumer goods companies — managing FX risk across 50+ markets, running intercompany settlement for $50B+ enterprises, and building the fintech infrastructure that modern global operations demand.

We didn't study this problem from the outside. We lived it: the month-end reconciliation crunches, the spreadsheets that nobody trusts but everyone depends on, the compliance documentation assembled by hand for every corridor. We built Trezh because we know exactly where the process breaks — and what it takes to fix it at scale.

62%
of new TMS evaluations
favor cloud deployment
12–18mo
average legacy TMS
implementation timeline
74%
of treasury teams
expanding AI use
$6.6B+
treasury management
software market

We're working with a small number of treasury teams to shape the platform.

If your organization manages intercompany flows across multiple entities and jurisdictions, we'd like to talk.