PhotonMatters
Books that

Books that balance themselves.

AI auto-reconciliation that matches payments, ledgers and gateway settlements at scale, flagging breaks before they age and turning month-end into a non-event.

Overview

Every rupee, dollar, naira, rand, kwacha and dirham accounted for

Lenders in emerging markets receive money through a tangle of channels: bank transfers, PSP gateways, mobile-money wallets and cash agents, each with its own file format, cut-off time and fee logic. Reconciliation AI ingests them all, matches each inflow to the right loan, instalment and ledger entry, and surfaces only the genuine exceptions.

What used to take a team of analysts days of spreadsheet work now runs continuously in the background. Breaks are detected the moment they appear, aged automatically, and routed to the right resolution workflow, so leakage is caught early and the close happens on time.

Near-zero manual matching

AI handles the long tail of partial, batched and out-of-order payments.

Faster month-end close

continuous reconciliation replaces the end-of-period scramble.

Fewer leakages

every break is aged, owned and resolved with a full audit trail.

Continuous, AI-driven reconciliation
Capabilities

Everything you need to reconcile at scale.

From raw multi-source ingestion to a board-ready audit trail: one engine for the whole reconciliation lifecycle.

01

Multi-source ingestion

Bank statements, core-ledger entries and PSP / mobile-money settlement files: any format, any cadence, normalised on the way in.

02

AI auto-matching

Models match one-to-one, one-to-many and many-to-many, handling fees, FX, partial and batched payments without rigid rules.

03

Exception & break detection

Unmatched, duplicated, short-paid and over-paid items are isolated instantly, so analysts only ever touch real discrepancies.

04

Break aging & resolution

Every break is aged into buckets, assigned an owner, and tracked through a configurable resolution workflow with SLAs.

05

Allocation & suspense

Payments are allocated to the correct loan and instalment; unidentified funds park in suspense and clear automatically when matched.

06

Audit trail & reporting

Immutable, time-stamped history of every match and adjustment, with reconciliation dashboards built for finance and auditors.

How it works

  1. 01

    Ingest

    Pull in bank statements, ledger entries and PSP / mobile-money settlements; normalise formats and currencies.

  2. 02

    Match

    AI auto-matches inflows to loans, instalments and ledger lines across complex many-to-many relationships.

  3. 03

    Flag exceptions

    Unmatched and discrepant items are isolated, aged and routed to the right owner with an SLA clock.

  4. 04

    Resolve

    Workflows clear breaks, post adjustments and update the ledger, leaving a complete, auditable trail.

Outcomes

Near-zero manual matching

analysts focus only on genuine exceptions instead of spreadsheets.

Faster month-end close

continuous reconciliation replaces the end-of-period scramble.

Fewer leakages

breaks are caught and aged early, before they become write-offs.

See it on your data

Reconcile a real settlement file with us.

Bring a day of your bank, PSP and mobile-money files. We’ll show you what auto-matches, what breaks, and how much faster your close could be.