Process Automation Service

There is a moment in every automation conversation — usually around slide fourteen of the vendor deck — where someone says the word "seamless." This is your cue to become suspicious.

Nothing involving software, people, legacy systems, and organisational politics is seamless. The honest word is "engineered." Engineered means someone sat down, understood the mess, chose the right tools for the specific problem, built something that handles the exceptions as well as the happy path, tested it against reality, and documented it so that the next person does not need a séance to understand what it does.

That is what process automation is, when done properly. Not a demo. Not a proof of concept that lives eternally in a sandbox, impressing people who will never use it. A working automation, in production, doing real work, with logs you can read and a plan for when things go sideways.

Here is what that involves.

What "Process Automation" Actually Means Here

The phrase "process automation" has become so capacious it now covers everything from a Zapier trigger to a full enterprise transformation program. This is not helpful. So here is what we mean, specifically.

You will receive one of the following solution types — or a blend, where the work demands it:

Workflow automation — Routing, approvals, SLAs, escalations, and audit trails. This is the orchestration layer: software that sits on top of your people and systems to manage handoffs and exceptions, so that work flows instead of pooling in someone's inbox like water behind a blocked drain. McKinsey calls this "smart workflow." We call it "the thing that stops work disappearing into the gap between two departments."

System integration (API / webhook / iPaaS) — Data moves automatically between your systems — CRM, ERP, ticketing, email, spreadsheets — with validation, retries, and error handling. This is plumbing. And like all plumbing, it is invisible when it works and catastrophic when it does not. Nobody has ever complimented their plumber at a dinner party, but everyone notices when the kitchen floods.

RPA (robotic process automation) — When the only interface is a legacy UI that pre-dates the concept of an API, RPA can automate rules-based tasks: moving data between email and systems, generating reports, checking files. We use it when there is genuinely no better option. We do not pretend it is a strategic platform. It is a workaround with a marketing budget, and we treat it accordingly.

AI-enabled automation — For the language-heavy steps that rigid rules cannot handle: triage, classification, extraction, summarisation, drafting, routing. Always paired with confidence thresholds, guardrails, and human review where the stakes warrant it. AI is excellent at reading messy inputs. It is not yet excellent at being held accountable for what it decides to do about them.

Lightweight web app or internal tool — A secure intake portal, an operations dashboard, an exception queue. Built only when the workflow genuinely needs a clean front-end rather than yet another workaround bolted onto a spreadsheet that has, somewhere along the way, become critical infrastructure without anyone formally acknowledging this.

The choice of which approach to use is an engineering decision, not a marketing one. The right answer is always the simplest thing that works reliably. Glamour is not a selection criterion.

What You Get

1. Solution Design Package

Every automation project begins with a design phase, for the same reason every building begins with an architect: because discovering structural problems after construction is significantly more expensive — and considerably more embarrassing — than discovering them before.

The design defines what is in scope, what is not, and what success looks like in terms that can be measured, not merely felt.

• Scope definition — what is in, what is out, and the success metrics that will tell you whether it worked; ambiguity at this stage is the single largest source of cost overruns, scope arguments, and mutual disappointment later

• To-be workflow — the happy path and the key exception paths; the happy path is what happens in demos; the exception paths are what happens in production, which is where your automation will actually live

• System and data mapping — where data comes from, where it goes, what fields matter, and what validations are needed; this is the anatomy of the integration, and getting it wrong is like wiring a house from memory

• Architecture diagram — components, integrations, authentication, logging, and storage; not for decoration, but so that anyone maintaining this in eighteen months can understand what it does without having to reconstruct your thinking from first principles

• Risk and controls review — access points, auditability, error handling, and compliance touchpoints; the things nobody wants to discuss in the design phase and everyone desperately wishes they had discussed once a regulator asks an awkward question

2. One Production-Ready Automation

This is the thing itself. Not a prototype. Not a demo. A complete automation, built to production standards, doing real work in your real environment.

• Core workflow logic — rules, routing, SLAs, retries; the decision-making engine that determines what happens, when, and what to do when the expected does not occur

• Connectors and integrations — API, webhook, or RPA, depending on what the systems involved actually support, as opposed to what their marketing materials optimistically claim they support

• Input validation and data hygiene — because the single most common cause of automation failure is not bad logic but bad data; garbage in, garbage out is not a cliché, it is a law of nature, and no amount of clever engineering overrides it

• Exception handling — a dead-letter queue or equivalent, so that when something fails — and something always fails — it fails visibly, into a place where a human can find it, understand it, and fix it, rather than vanishing silently into the void

• Secure credential handling — no passwords in scripts, no API keys in plain text, no "we'll fix the security later" optimism that never survives contact with an auditor

• Environment setup — a dev/test/prod pattern where feasible, because testing in production is not a strategy, it is a confession

3. Testing and Quality Gates

The purpose of testing is to find out where your automation breaks before your customers do. This sounds obvious. It is also, in practice, the phase most likely to be compressed, skipped, or politely described as "we'll do a quick smoke test."

We do not do a quick smoke test.

• Test plan — covering the happy path, the edge cases, and the failure modes; an automation that only works when everything goes perfectly is not automation, it is a coincidence

• UAT support — guided testing with your subject-matter experts, because the people who know the edge cases are the people who live with them every day, not the people who designed the flowchart

• Performance checks — where volume warrants it; an automation that works beautifully for ten cases and collapses at a thousand is a prototype wearing a production badge

• Release checklist and rollback approach — because knowing how to undo what you just did is as important as knowing how to do it, and "ship it and pray" is not a deployment strategy

4. Handover Documentation

Documentation is the thing that separates a professional deliverable from a hostage situation. Without it, your automation works only as long as the person who built it is available, willing, and not on holiday.

• Runbook — how it works, where to look when it fails, and the common fixes; written for the person who inherits this at 9am on a Monday with no context, not the person who built it

• Support playbook — alerts, escalation paths, and ownership assumptions; who gets paged, in what order, and what they should do first

• User SOP — one to two pages explaining what changes for the team; the fastest way to undermine an automation is to surprise the people it affects

• Configuration reference — variables, credentials, schedules, endpoints; the settings that someone will inevitably need to change six months from now, documented in a place they can actually find

5. Monitoring Basics

An automation without monitoring is a lodger without house rules. It might be fine. It might also be quietly making the same mistake ten thousand times a day while the dashboard shows green — because the dashboard measures activity, not correctness.

• Event and error logging — for key steps, so you can reconstruct what happened and why, rather than staring at a blank screen muttering "it was working yesterday"

• Health and throughput view — a dashboard or report showing whether the automation is doing what it should, at the volume it should, with the success rate it should

• Alerting hooks — email, Teams, Slack, or whatever your organisation uses; the alert should say "the right outcome did not happen," not merely "the bot ran"

You will also receive OCM-ready materials — an SOP, a workflow summary, and a rollout checklist — so that the people affected by the change have something to read that was not written by a developer in a hurry. Change management execution remains with the business, because we can hand you the map but we cannot make people walk the route.

What Is Not Included (Unless You Add It)

Honesty about scope is more useful than ambiguity, so here is what this package does not cover:

Full organisational change management — communications plans, training delivery, role redesign. We provide the materials. You provide the organisational momentum. These are not the same skill set, and pretending otherwise helps nobody.

Multi-process programmes or centre-of-excellence build-outs — this is one automation, done well. If you need a program, we are happy to discuss one. But a program that starts with a single well-executed pilot is worth more than a program that starts with a strategy deck and never ships anything.

Multiple automations — this package delivers one. One properly built, properly tested, properly documented automation is worth more than three that sort of work.

24/7 support or ongoing managed services — we hand over a working automation with documentation and monitoring. We do not move in permanently, though we are available if you need us.

Licensing costs for platforms or tools — your organisation typically provides these. We advise on what is needed during the design phase so there are no surprises.

Pricing

Pricing in consulting is usually an exercise in strategic ambiguity. Ours is not. There are three tiers, they reflect genuine differences in complexity, and none of them contains a hidden "phase two" that doubles the cost.

Starter Automation — $14,900

Good for one workflow with clear rules and low exception rates. The kind of process where someone says "I do the same thing forty times a day" and means it literally.

• One trigger connected to one or two systems

• Minimal data transformation

• Basic exception handling and notifications

• Single environment deployment

This is the automation equivalent of a well-fitted kitchen shelf. Not glamorous. Entirely functional. Solves exactly the problem it was designed to solve, and nothing more.

Standard Automation — $19,900

Good for most real business processes — the ones that involve multiple systems, genuine branching logic, and the kind of exceptions that make simple rule-based approaches quietly fall apart.

• Two to four system integrations (CRM, ERP, email, forms, ticketing)

• Substantive validation and branching logic

• Exception queue with retries

• Full UAT cycle and production rollout

• Solid runbook and monitoring hooks

This is where most organisations continue after a process audit. It handles the gap between "this process looks simple on a whiteboard" and "this process involves four systems, three approval levels, and a spreadsheet that Karen maintains and nobody else understands."

Complex Automation — $49,900+

Good for high exception rates, brittle legacy systems, heavier compliance requirements, or situations where the automation needs a small application built alongside it.

• RPA combined with integrations, or a custom web UI, or both

• Multiple roles, approvals, and audit trails

• Advanced AI decisioning with guardrails

• Broader testing surface and operational hardening

• Includes a full process audit

This is for the processes that have defeated previous automation attempts. The ones with enough edge cases, legacy dependencies, and organisational scar tissue that they require genuinely careful engineering rather than a template and optimism.

Everything Else — $290/hr

Extensions, training, team enablement, ongoing support, dogsitting or anything that falls outside the three tiers. One flat rate. No surprises. No "we'll need to scope that separately" conversations that end in a number three times what you expected.

The Honest Version

Here is what we are actually selling: one automation, done properly, that works in production, handles exceptions, has documentation a human can read, and can be maintained by someone who was not in the room when it was built.

That sounds simple. It is not. The reason most automations disappoint is not that the technology failed. It is that someone skipped the design, compressed the testing, forgot the exceptions, and documented nothing. Then they called it "phase one" and moved on, leaving behind an automation that works only in the sense that it runs — not in the sense that it produces the right outcome.

We do not do that. We do the tedious, unglamorous, essential work of building something that actually works, and then we hand it over in a state where you can operate it, maintain it, and explain it to an auditor without needing to phone us first.

Which, when you think about it, is the only honest definition of "done."

Book a discovery call

If you are not sure which tier fits, or whether your process is ready for automation at all, start with a conversation. We will tell you what is worth automating, what needs process work first, and what you should leave alone entirely.

Learn more: process audit | Learn more: process automation | Contact us