Know the scope.
Know the price.

Only when we need to understand the workflow, data, or exceptions before quoting responsibly. If you already have a clear scope, we can go straight to an implementation proposal. A focused working session may be enough for one technical decision.

It's a separate $2,500 engagement, but if you go ahead with a build within 60 days, half the audit fee is credited against it. The audit covers one workflow, its bottlenecks and constraints, and a written recommendation with an implementation scope and budget. You keep the plan either way. A build starts at $7,500.

One narrow workflow or focused tool with agreed acceptance criteria, testing, documentation, and handover. For example, a repeatable intake-and-routing process using an existing data source. Multiple integrations, historical data migration, sensitive data, and larger interfaces need their own scope. The budget guide is a planning aid; the written proposal sets the price.

We discuss the change, its effect on cost and timing, and your options before adding the work. The written proposal sets out the deliverables, acceptance criteria, payment milestones, and responsibilities.

Care is optional and covers a system Bootable delivered. It includes monitoring, a monthly health check, updates when connected tools change their APIs, and fixes for defects in the delivered work, with an agreed response window. New features and improvements are quoted separately or billed at the advisory rate with a spending cap. It does not include round-the-clock incident response.

A starter automation usually takes about two weeks. A focused build typically takes three to six weeks, depending mostly on how quickly we can get access to your tools and feedback from the people doing the work. The proposal states the schedule and what it depends on.

You do. Once each invoice is paid, the code, configuration, and documentation delivered for your project are yours. Automations run in accounts your business owns, not ours, so you are never locked in. We may reuse general techniques and non-confidential building blocks on other projects.

Usually inside tools you already pay for, or on a small cloud service in your name. AI model usage is billed by the provider based on volume, so we estimate it during scoping and add spending limits. For most small-business systems, expect hosting, subscription, and model costs between nothing and a few hundred dollars a month. The proposal lists them before you commit, and they're billed at cost.

Fixed-fee work is invoiced in milestones, typically half to start and the balance on acceptance; larger projects use more milestones. Working sessions and audits are paid up front. Care is billed monthly and can be cancelled with 30 days' notice.

Everything is built to be handed over: documented, running in your accounts, and using mainstream tools any competent developer can pick up. The handover includes a runbook covering what runs where, how to check it, and how to change it. Ongoing care is there if you'd rather we keep looking after it.

No. We start with the process. AI agents and RAG assistants are the right fit when the work involves reading, judging, or writing language. If a direct integration or a simple automation does the job better, that's what we'll recommend.