Our System · The onsoftarch Approach

Don't replace what already works. Build around it.

Your legacy system isn't a problem to be solved. It's decades of proven logic running your business every day, and it's still here for a reason. Our approach respects that.

We leave your core untouched and build a modern layer around it: APIs, dashboards, health monitoring, and integrations.

Talk to Our Team Watch the transformation

Our Philosophy

"If a system has run reliably for thirty years, the system isn't the risk. Replacing it is." Legacy platforms survive because they work. They've absorbed decades of edge cases and real-world stress no rebuild can replicate. We treat that reliability as an asset to protect.
01

Proven beats new

A system that's processed billions without failure has earned its place. Decades of stability shouldn't be traded for novelty.

02

The risk lives in the rewrite

Replacement projects are where outages and overruns happen. The safest path forward rarely means starting over.

03

Modern is a layer, not a replacement

You don't need a new core for modern capability. You need the right architecture built around the one you trust.

The Architecture

Your core, preserved.
Everything else, modern.

Our approach is layered. The proven system stays exactly as it is at the centre. Around it, we build the modern capabilities you need, each connected through clean interfaces.

Outer Layer · Business Value

Dashboards & Decision Tools

Real-time dashboards that turn your legacy data into decisions, tailored to each team.

↑ connected via clean APIs ↑
Middle Layer · Integration

API & Integration Fabric

A modern fabric that lets your legacy core talk to cloud services, apps, and external systems in real time.

↑ wrapped without modification ↑
Core · Untouched

Your Proven Legacy System

The COBOL, PL/I, or mainframe logic that's run your business for decades. Not rewritten. Not migrated. Not at risk.

From Code to Clarity

The same system. Suddenly you can read it.

Watch a wall of legacy COBOL become a live, decision-ready dashboard, without a single line of the original code changing.

PAYSWCH.CBL · core ledger · unmodified
* IDENTIFICATION DIVISION.
* PROGRAM-ID. PAYSWCH.
DATA DIVISION.
WORKING-STORAGE SECTION.
01 WS-TXN-COUNT PIC 9(9) VALUE ZERO.
01 WS-TXN-TODAY PIC 9(9) COMP-3.
01 WS-VOL-PROCESSED PIC 9(11)V99 COMP-3.
01 WS-CURR-CODE PIC X(03) VALUE 'ANG'.
01 WS-AUTH-OK PIC 9(7).
01 WS-AUTH-DECLINE PIC 9(7).
01 WS-AUTH-RATE PIC 9(3)V9(2).
01 WS-CARD-TABLE.
05 WS-VISA-AMT PIC 9(11)V99 COMP-3.
05 WS-MC-AMT PIC 9(11)V99 COMP-3.
05 WS-APAY-AMT PIC 9(11)V99 COMP-3.
05 WS-GPAY-AMT PIC 9(11)V99 COMP-3.
PROCEDURE DIVISION.
0000-MAIN-PROCESS.
ADD 1 TO WS-TXN-COUNT.
COMPUTE WS-VOL-PROCESSED = WS-VOL-PROCESSED + TXN-AMT.
IF AUTH-RESPONSE = '00'
ADD 1 TO WS-AUTH-OK
ELSE
ADD 1 TO WS-AUTH-DECLINE
END-IF.
EVALUATE CARD-SCHEME
WHEN 'VI' ADD TXN-AMT TO WS-VISA-AMT
WHEN 'MC' ADD TXN-AMT TO WS-MC-AMT
END-EVALUATE.
PERFORM 9000-WRITE-LEDGER.
0000-EXIT. GOBACK.
Transactions Today
0
↑ 12% vs last week
Volume Processed
ANG 0
↑ 8% vs yesterday
Auth Success Rate
0%
Payments by Card Type & Wallet
VisaANG 743K · 40.4%
MastercardANG 581K · 31.6%
Apple PayANG 226K · 12.3%
Google PayANG 121K · 6.6%
Throughput · 30 min
4.2 TPS · peak 11.8
A wall of working-storage and procedure logic. Reliable, but unreadable.

What We Build Around It

Modern capability,
zero core risk.

Four pillars, each adding modern value without touching a line of the logic you depend on.

API Layer

Integration Through Clean APIs

We expose your system's data and functions through clean, documented APIs. The platform that couldn't talk to anything now connects to cloud services, apps, and analytics, all without changing the core.

The island becomes part of the mainland, at your pace.

Dashboards

Decisions, Not Just Data

We build live dashboards that surface what your system already knows: volume, trends, risk, and performance, tailored to the people making decisions.

Decades of data, finally working for the business.

Health

Continuous Health Monitoring

We instrument your system to watch its own vital signs: throughput, latency, error rates, and capacity. You get early warning before problems become outages.

Know it's healthy. Know before it isn't.

Integration

Connecting Old & New

We make your legacy core a first-class citizen of your modern stack. New apps, cloud platforms, and third-party systems integrate cleanly, with the core never at risk.

Build the future without rebuilding the foundation.

Our Software

ON/2

Customisable COBOL · Fault-Tolerant by Design

ON/2 is our homegrown, fully customisable COBOL platform, built for workloads that cannot fail. It runs on our fault-tolerant servers, engineered to process high-volume payments without dropping a transaction.

It's in production at banks today, processing real payment traffic, and has grown alongside those businesses as their needs changed.

ON2-CORE · node status
* ON/2 fault-tolerant cluster

PRIMARY-NODE : ACTIVE ✓
FAILOVER-NODE : STANDBY ✓
REPLICATION : IN-SYNC ✓
UPTIME : 99.999%
PEAK-THROUGHPUT: 3,200 TPS
FAILED-TXNS : 0

* customised per institution
* zero downtime in failover
Customisable

Shaped to your institution

We adapt ON/2's COBOL core to each bank's products, rules, and workflows, then keep evolving it as the business grows.

Fault-Tolerant

Built to never stop

Active-standby clustering, live replication, and automatic failover mean a node can fail without losing a single transaction. Built for five-nines uptime.

High Volume

Payments at scale

Built to process high-volume payment traffic under pressure, the exact conditions a bank's switch faces at peak.

In production
Live at multiple banks, processing real payment traffic today.
Battle-tested
Well used and well liked by the teams who depend on it daily.
Grows with you
Adapted and extended as our clients' businesses have scaled.

"The same fault-tolerant foundation that runs a bank's payment switch can underpin any application where downtime simply isn't an option."

Because its strength is reliability, ON/2 reaches beyond payments. Any high-uptime, high-stakes application can run on the same fault-tolerant servers banks already trust.

Built For Every Team

One system.
Insight for everyone.

We surface the right view for each team, so one trusted core drives decisions across the whole organisation.

Executive

Growth & risk at a glance

Volume, unit economics, and the few numbers leadership needs to steer.

Finance

Margin & settlement clarity

Revenue, fees, and exposure in real time, not waiting on month-end batches.

Technology

Capacity & reliability

Latency, throughput, and capacity, so engineering can plan ahead instead of reacting.

Operations

Live operational view

The day-to-day signals teams need to keep things running.

Observability

A legacy system you
can finally see into.

Most legacy platforms are black boxes: reliable, but opaque. Our monitoring layer surfaces real health in real time, so you can trust it with evidence and act early.

No core changes. Just visibility where there used to be none.

* onsoftarch health monitor · live

CORE-STATUS : OPERATIONAL ✓
UPTIME : 99.999%
THROUGHPUT : 4.2 TPS (68% cap)
AUTH-LATENCY (p95) : 134 ms
BATCH-WINDOW : WITHIN LIMITS ✓
ERROR-RATE : 0.02%
CAPACITY-HEADROOM : HEALTHY ✓

* core untouched · observed externally

Security & Compliance

Built for the standards
banks are held to.

We work inside regulated payment environments every day. Security and compliance aren't features we bolt on, they shape how we design, deploy, and operate everything we build around your core.

PCI-DSS

Payment-grade controls

Our integration and dashboard layers are designed to operate within PCI-DSS scope, with cardholder data handled, masked, and logged to the standards your auditors expect.

Data Handling

Least-privilege by default

We minimise data exposure: the wrapper reads only what a given view needs, sensitive fields stay masked, and access is scoped, authenticated, and fully audit-logged.

Deployment

Inside your perimeter

Our systems run where your core runs, on-premise or in your private cloud, behind your existing controls. Nothing critical is forced out to third-party infrastructure.

Auditability

Evidence, not assertions

Every transaction the layer touches is traceable. We add the logging, monitoring, and reporting that turn a regulator's questions into a query, not a fire drill.

Continuity

No lock-in, no cliff edge

Source ownership, documentation, and optional code escrow mean you are never dependent on a single vendor's survival to keep your system running.

Change Control

Reversible, governed change

Because the core is never modified, every change we make is isolated, testable, and reversible, fitting cleanly into your existing change-management process.

PCI-DSS aligned Least-privilege access Full audit trail On-prem / private cloud Code escrow available Reversible change control

Keep the system.
Gain everything around it.

See what our approach looks like on your platform, with the core you trust kept exactly as it is.

Book a Technical Review