Claude Code Setup + Training for Your AU Team

Most teams that “tried Claude Code” stalled because the install was clean but the workflows weren’t. We solve the workflow piece. The install is the easy part.

What Claude Code actually does for a non-technical team

Claude Code is Anthropic’s command-line version of Claude. The “code” in the name scares people off, but the most valuable workflows for an ops team have nothing to do with programming. The difference from Claude in a browser tab is that Claude Code can read and write the actual files on your machine, run through a folder of documents, and follow a saved set of instructions every time. It is the difference between an assistant you brief from scratch each morning and one that already knows your project and can touch your files.

For an AU services business that means jobs like these become a single command instead of an afternoon:

• Bulk document work. Rename, reformat or restructure a folder of 200 files. Pull the key fields out of 50 supplier PDFs into one spreadsheet. Convert a pile of messy notes into clean SOPs in your template.
• Repeatable reports. A weekly client status report that reads from the same sources, in the same format, written to the same standard, every Monday, because the instructions live in the project rather than in someone’s memory.
• SOP and content QA. Run every SOP against a checklist for gaps, outdated steps or missing screenshots. Check a batch of blog drafts against your style guide.
• Image and file processing at volume. The kind of repetitive batch work a VA otherwise grinds through by hand.

The unlock is the project workspace and the CLAUDE.md file. CLAUDE.md is a plain-English instructions document that sits in the project and tells Claude Code how your business works, your tone, your rules, your file structure, so it behaves consistently for whoever runs it. Set that up well and your VA gets your standards baked in; skip it and you get a clever tool nobody trusts.

Why teams stall, and what we fix

Plenty of teams install Claude Code and then quietly stop using it. It is almost never the install. It is that nobody mapped which of their actual workflows it should own, nobody wrote the project instructions, and the team was left to figure out commands on a busy week. So it stays a curiosity instead of becoming the thing that does Monday’s reports.

Our engagement is built around that gap. We run a 90-minute discovery to map your current workflows and pick the ten with the highest payoff, the admin and ops work that scales linearly with your volume and is eating your team’s week. We install and configure Claude Code on the team’s machines, set up project workspaces with your CLAUDE.md, SOPs and templates, and build those ten workflows as ready-to-run templates for your business, not generic examples.

Then we train. A live 90-minute session for the founder, VA and ops team, plus ten recorded 3-to-5-minute walkthroughs so a new hire can learn each workflow without booking your time. You get a 30-day usage-support Slack channel and a performance baseline with a 30-day check-in, because the measure of success is whether it is still in daily use a month later, not whether the demo went well.

How the engagement runs

It is a fixed-scope project, $2,000 to $5,000 depending on team size, over 2 to 3 weeks. What we need from you is access to the team’s machines for setup, your existing SOPs and templates, and the founder plus VA plus ops in the room for discovery and training. The ROI is strongest with a 3-to-5 person team where the same admin work repeats across people; a single VA can benefit, but a solo operator is usually better served by our Claude Code playbook guide than a full team engagement.

Where it fits, and where it breaks

Claude Code is excellent at structured, repeatable file-and-document work where the instructions can be written down once and reused. It is not a fit for everything. Tasks that need a person’s judgement on every instance, a sensitive client conversation, a one-off decision with real money on it, are still human work. It can make mistakes on a task, which is why the workflows we build keep a person reviewing anything that goes out or commits the business. And like any tool, an unused setup decays: the project instructions need the occasional update as your business changes, which is why the handover and check-in matter.

A few honest questions people ask

Do team members need to be technical? No. Claude Code runs in a command line, but the workflows we train do not require coding. A VA is typically productive within the first training session.

What does it cost to run after the project? Claude Pro covers most usage; heavy users may move to Claude Max. We sort out the right subscription strategy in scoping so you are not over- or under-paying.

Can it work for a single VA? Yes, but the return is stronger across a 3-to-5 person team. For solo VAs, start with our Claude Code for non-developers guide.

The ten workflows, and how we choose them

We do not arrive with a generic list. The ten workflows come out of your discovery session, scored on two things: how much time the task eats now, and how repeatable it is. The best candidates are the jobs that happen weekly or daily, follow the same shape every time, and currently scale linearly with your volume, which is to say they get worse as you grow. That is where a saved, instruction-driven workflow turns a recurring afternoon into a recurring few minutes.

The strongest fit is an AU services business that already has a VA and an ops function, where roughly 60% of the work is admin and ops that repeats across people. A founder who personally uses Claude Pro but whose team has not made the jump to Claude Code is the classic case: the leverage is in getting the team productive, not the founder. If you have no team to train, the playbook guide is the better starting point than a full engagement, and we will tell you that before you spend.

What “done” looks like

Every DotVA AI services engagement is fixed-scope, fixed-price, and fixed-timeline. No retainers without a defined deliverable. No “discovery phases” that turn into 6-month projects. The scoping call is free; the written quote that follows is what you sign off on.

Why we say no

We turn down roughly 30% of applications. Usually because the team isn’t ready, the goal isn’t clear, or there’s a simpler path that doesn’t need us. We’d rather tell you that on the discovery call than take a brief that won’t deliver. The shortest path to outcome isn’t always us.

Apply below

Two minutes. We read every application within 1 business day. If we’re a good fit, we book the scoping call. If we’re not, we say so – and usually suggest who is.