What AI automation actually costs

Three ways to work with us, what each one includes, and an honest account of what moves the number. The discovery call is free, and you leave it with a scope either way.

Engagement models

Pilot / Sprint

You want proof before you commit. A fixed-scope build of two to four weeks, aimed at one process. Deliverables and success criteria are written down before any work starts, so there is no argument at the end about what 'done' meant.

• One process mapped end to end
• One working agent or n8n workflow, deployed
• Integrations with the tools you already use
• Handover docs and a recorded walkthrough

Monthly retainer

The automation is live and now has to stay alive. Ongoing operation and improvement of what is running: monitoring, fixing what breaks when an upstream tool changes, and shipping the next workflow in the queue.

• Monitoring and incident response
• A standing queue of new workflows
• Prompt and model updates as they change
• A named person who knows your setup

Custom / Enterprise

Several systems, several teams, or a compliance review in the way. Scoped from a discovery phase rather than a price list. Typically where data residency, self-hosted n8n, audit trails or procurement are part of the conversation.

• Discovery phase with written findings
• Self-hosted n8n where data cannot leave
• Security review and DPA
• Phased delivery against milestones

What moves the price

• How many processes you want automated, and whether they share the same systems.
• How many integrations are involved, and whether any of them lack a usable API.
• Data volume, and whether the work can run on a schedule or has to be real-time.
• Compliance: self-hosting, data residency, audit trails and review cycles all add time.
• How clean the current process is. An undocumented process has to be mapped before it can be automated.
• Who maintains it afterwards — handover to your team, or we keep operating it.

Pricing questions

Why are there no fixed prices on this page?

Because the honest number depends on scope, and we would rather not publish one we have to walk back on the call. What is fixed is the model: a pilot is a fixed-scope build agreed in writing before it starts, so the price you agree is the price you pay.

Do you charge per agent?

No. We price the work, not the number of agents. One well-built agent covering three jobs costs less to run and maintain than three thin ones, and we will say so if that is the better design for your case.

What if the pilot does not deliver?

The pilot has written success criteria agreed before the build. If we do not ship working automation against them, the milestone is not payable. The exact contractual terms are being finalised.

Is there a minimum engagement?

Retainers carry a minimum term so the work has time to compound rather than being abandoned mid-build; pilots have no minimum beyond their own scope.

What does it cost to run once it is built?

Two separate things: our fee, and your own running costs. Running costs are the model API usage and, if you self-host, the server. We size both during discovery so you see the monthly figure before you commit, not after.

Do you work with businesses outside Pakistan?

Yes. We work with clients in the United States, United Kingdom and the Gulf, and book calls across those working hours. Delivery is remote, with a named person on your account rather than a rotating queue.

Can we start smaller than a pilot?

The discovery call is the smaller start, and it is free. Thirty minutes is usually enough to tell whether a process is worth automating at all, which is the question most worth answering before any money moves.

How do you handle payment?

Pilots are billed against the agreed milestones and retainers are billed monthly.

Book a discovery call