Services

What working with us costs

The written assessment is free. The audit is the only fixed price on this page. Everything after that is quoted in a written scope before work starts. The work covers recurring operating workflows such as intake, routing, follow-up, reporting, documents, and inbox triage.

Freewritten, async

Free written assessment

Deliverable. We identify the step causing the trouble and recommend one fix. Sometimes that fix is a tool you can buy without us.

Delivery. A written reply by email after we read the form.

You provide. The form, an honest description of what keeps getting stuck, and an optional real example with anything sensitive redacted.

Free is deliberate. It routes you to the smallest useful next step, including the one where you take the answer and never hire us.

$1,000 to $5,000one time

AI Workflow Audit

Deliverable. A workflow map, ranked backlog, 90-day plan, honest hours-saved math, and the first sprint we would run.

When it fits. You know the work is costing you and you cannot yet say which workflow, or where the real break is.

Start condition. The working session, required access, and source examples need to be in place before the audit begins.

You provide. Someone who knows the workflow, access to the relevant systems or exports, and enough real examples to map the work as it actually runs.

Quotedfixed price, in writing

AI Ops Sprint

Deliverable. One working workflow built end to end, with the system, specs, human approval gate, SOP, and measurement plan. Fixed price, set in the written scope before the build starts.

When it fits. One workflow, already understood, with edges you can describe. You do not need an audit first to get here.

Delivery. The written scope sets the build stages, review gates, and acceptance condition.

You provide. System access, a decision-maker for approval gates, real inputs to test, and timely review while the workflow is being built.

See the sprint structure

Monthlyquoted in writing

AI Operating Partner

Deliverable. One or two workflow improvements a month, documented and yours. We stand up the operating layer one useful system at a time.

When it fits. More than one build is worth doing and you want them shipped in order, without hiring for it.

Structure. Month to month, no lock-in, priced in a written scope we review as the backlog moves.

You provide. A clear priority order, access to the systems in scope, and a client-side owner who can review decisions and keep the work moving.

See the operating-partner structure

You work directly with the person running your engagement. We agree on the scope and schedule before work starts.

Which engagement you actually need

If you can already describe the workflow and its edges, we scope a sprint and start. If the problem is broad, or you cannot yet name which workflow is costing you the most, the audit answers that first.

Start here

Free written assessment

Need the map

AI Workflow Audit

The drag is real, but the first build is not obvious. Map it, rank it, and leave with the plan.

Know the workflow

AI Ops Sprint

One defined workflow needs to work now, with an approval gate, documentation, and a clean handoff. Skip the audit and scope the build.

Have a backlog

AI Operating Partner

There is more than one useful build and you want steady implementation without creating an internal bench.

Free assessment or paid audit
The free assessment

One email about one process. You describe how it runs today, we reply with where it gets stuck, one change we would make, and whether a tool you already have could do it. You can stop there. See what the reply looks like.

The paid audit

A working session and the full map: every step of the workflow traced as it runs, a ranked backlog, the hours-saved arithmetic with its assumptions shown, and the first sprint we would run. You can hand that plan to anyone, including someone other than us. What happens inside an audit.

How they relate

The assessment is where you find out whether you need the audit at all. The audit fee credits in full toward a sprint booked within 90 days.

What moves the price, and what you pay separately

A range exists because these five things differ by company. The written scope prices your version of them before work starts.

Price drivers Why a scope grows
Systems in the path

Every system the workflow has to read from or write to adds an integration, a credential, and a failure mode somebody has to handle.

Exceptions

A workflow with three clean paths costs less than one with thirty special cases. The exceptions are usually where the work is.

State of the data

If the same record lives in three places with three spellings, part of the build is reconciling that before anything can route on it.

Review load

Every customer-facing step needs a reviewer, a queue, and a defined path for what happens when the reviewer says no.

Access and review on your side

Test data, sandbox access, and your own security or legal review take calendar time. The schedule in the scope accounts for it.

Third-party costs Yours, not ours
Software and API charges

Model providers, automation platforms, and the SaaS tools a workflow runs on are billed to your accounts at their own rates. Those charges sit outside our fee and stay yours.

Usage moves with volume

Model and platform charges scale with how often the workflow runs. The scope names the tools a build needs, so you can price them against your own volumes before you agree to it.

Vendors change their prices

They do it on their own schedule and we do not control it. Section 10 of the terms says the same thing in the language that binds.

Ownership and trust

The handoff, and what happens after it

The handoff has to stand on its own. Here is what lands in your hands and where our work stops.

What the handoff contains Every build
Accounts and access

The workflow runs on accounts in your name. You hold the logins and the billing, and we work under access you can revoke without breaking the system.

Documentation

Written instructions for whoever runs it: what each step does, what the approval queue expects of a reviewer, and the decisions the workflow is not allowed to make.

An export path

Where your data sits and how to get it out in an ordinary format. The handoff explains how to export your data in an ordinary format.

Failure notice

What happens when a step breaks: where the run stops, who gets told, and what the workflow will not send while it is stopped.

Recovery instructions

How to restart a stalled run, how to reprocess what it skipped, and how to do the work by hand while the workflow is down.

Vendor-neutral decisions

We pick platforms on fit, cost, and how cleanly you can leave them. The controls around your data are on the data handling page.

Support after handoff Stated limits
What is included

If a deliverable materially fails to match its written scope, tell us within 30 days of delivery and we re-perform the deficient part. That is the remedy in Section 11 of the terms, and it is the whole of what comes free after delivery.

How changes get scoped

Anything outside the written scope gets its own written scope, price, and schedule before it starts. A one-off change is usually a fixed quote. A steady stream of them is what the operating partner engagement is for.

What you maintain

Your accounts, subscriptions, credentials, and your own backups. When a vendor changes an API or retires a plan, that change lands on your side first and the runbook tells your team what it touches.

No response-time promise

There is no standard support response window. Any support coverage beyond the remedy above has to be agreed in the written engagement, so if you need a guaranteed response time, say so before we start.

What the written terms give you

The risk reversal In writing
Audit fee credit

The paid AI Workflow Audit fee credits in full toward a build sprint booked within 90 days of the audit.

Re-perform remedy

If a deliverable materially fails to match its written scope, tell us within 30 days of delivery and we will re-perform the deficient work. That is the remedy for defective work stated in Section 11 of the terms.

What you own

The systems, prompts, documentation, and reports built for you are yours once the engagement is paid in full. Section 7 of the terms covers the pre-existing tooling you get a license to keep using.

Named workflow products All six workflows
How a build runs

The working sequence, the approval gates, and what you see at each stage are on the process page.

The alternatives

How this compares

These are different ways to solve the same operating drag. The right answer depends on what you need after the build.

Hiring an employee

A sprint builds a defined system while your team keeps the judgment. An employee may be the better choice when the job needs daily context across changing responsibilities.

DIY tools

Tools still need someone to wire them together and maintain them. We build the workflow, put the risky output behind an approval gate, and hand it off documented. DIY is cheaper when someone on your team has the time and ownership to do that work.

A full agency

You work directly with the person building it, with no account layer in between. A full agency may fit better when you need several disciplines under one roof at once.

Start with the process that keeps getting stuck.

The free written assessment names the first fix and says which engagement fits, including when none of them does.

Get a free written assessment