The Processor

The engine that runs your digitised processes.

A production-grade automation runtime. Event-driven, retry-aware, fully audited. Reads signed events from your systems, executes the right actions, writes a complete trail of every step. Included with every engagement that digitises a process — and bound, by design, to the specification it runs.


01 / What It Does
What it does

One runtime. Every workflow.

The Processor is the layer between your systems — CRM, billing, product backend, internal databases, messaging — and the documented processes you want them to follow.

Absorbs the work spikes

When 80 customers all hit a renewal date the same week, the Processor queues every renewal, processes them through workers in parallel, and your team never sees the surge.

Never loses work

A third-party API hiccup doesn’t lose a job — it gets retried, automatically, until it succeeds or escalates to a human. Audit logs prove every step.

Tells you before customers do

A separate watchdog monitors queue depth, failure rates, and downstream API health. You hear about a problem on Friday morning, not from a customer support ticket on Monday.


02 / Architecture
Architecture

Event-in. Action-out.

The Processor is built around the same architectural pattern that powers high-throughput payment systems and message brokers. Events come in from anywhere; actions go out to everywhere; in between, a queue and a worker fleet absorb spikes, isolate failures, and never lose work.

EVENT SOURCES Your CRM Signed webhooks · real-time Your internal services Encrypted job requests Scheduled triggers Daily · monthly · quarterly QUEUE Signed · deduplicated Nothing dropped WORKER FLEET Worker 1 — service A Worker 2 — service B Worker 3 — service C … scales horizontally ACTIONS Your billing Create · update · sync Your CRM Mirror · enrich Notifications Email · alerts AUDIT LOG · OBSERVABILITY · RETRY · ALERTING Every job’s input, output, and timing recorded · Health monitor watches the runtime · Operations are zero-downtime
The Processor receives events, queues them, executes them through workers, and writes a full audit trail of every step.

03 / How Services Get Built
What you actually own — and why it changes the conversation

One Markdown file generates the running system.

Every digitised process you receive from NordOps is one Markdown file in your source repository. That file isn’t documentation alongside a running system — it generates the running system. Three things follow that change how you operate:

01 · Specification Process Markdown

One .md file per process. Steps, decisions, integration points, business rules, explicit out-of-scope clauses. The single source of truth.

02 · Generation Claude Code

Reads the .md as its design brief. Writes the service that implements it — constrained by what the Markdown says it can and cannot do.

03 · Runtime Service in the Processor

Plugs into the Processor and inherits queue, retries, audit logging, observability, and alerts. Bound by the .md.

If a future request needs behaviour the Markdown does not describe, the service does not silently extend itself. The Markdown is updated first; the service is regenerated to match. The two cannot drift apart.


04 / The Architectural Difference
The architectural difference

A development agency delivers code. NordOps delivers a constrained system.

An agency delivers code that is allowed to do whatever the developer thought sensible at the time. Six months later, no one quite knows what the system does, and any audit of “is this following our policy?” becomes a forensic exercise.

The Processor delivers a service that is constitutionally bound to its Markdown specification. The audit answer is the .md file. Always. Process changes are not engineering tickets handled in private — they are Markdown edits, version-controlled, reviewable in a pull request, and re-buildable by any AI coding agent including the buyer’s own.

What you get

  • Documentation that cannot drift from the system — because it generates the system
  • Audit answers in one Markdown file, current by definition
  • Changes via spec edit + regeneration — not via developer tickets
  • Full source ownership at handoff — Processor, services, Markdown, infrastructure
  • 60-day build warranty included on the entire delivery

05 / What You Own
What you own at handoff

Every layer. Forever.

The Processor is included with every NordOps engagement that digitises a process. It is not a separate purchase or a hosted service. At final handoff, the source code transfers to your repository and you own it — runtime, services, Markdown specifications, and infrastructure manifests — outright.

Runtime source code

The Processor itself. Modern JavaScript, Node.js LTS. Containerised for any infrastructure.

All deployed services

One service per digitised process, generated from one Markdown file. Plugged into the runtime, ready to run.

The Markdown specifications

The source of truth for every digitised process. Versioned in your repository, owned by you.

Infrastructure manifests

Docker, Compose, environment templates, deployment scripts. Reproducible from a clean checkout.

The Admin Portal

Renders Process Library views, monitors service status, manages access. Standard with every engagement.

60-day build warranty

From final handoff. Bug fixes and critical vulnerability patches at no charge. Optional maintenance picks up afterwards.

See the Processor running live.

A 30-minute discovery call includes a walkthrough of the Processor at the reference deployment and a first read on which processes in your business would benefit from being digitised.

Book a 30-minute call