Which tools can agents work in?

Anything with an interface, an import or a browser: Microsoft 365 and Google Workspace, HubSpot, Pipedrive and most CRMs, DATEV, Xero, QuickBooks, Slack, Teams, your ERP. We check the specific stack before the quote.

How is this different from Zapier or Make?

Those tools move data between fields when a trigger fires. An agent reads the content, makes a judgment within the rules you set, drafts text and asks a person when unsure. Many of our builds use both: the plumbing from a tool like that, the reading and writing from an agent.

What does the first workflow cost?

A fixed price, quoted after the call and depending on the workflow and your systems. Every quote comes in writing with milestones. Founding-client terms exist for the first cohort.

You can. Everything is handed over with documentation, and your IT partner or an in-house person can run it. Most clients keep a monthly retainer with us for monitoring and changes.

Workflow automation · the hours that repeat

The hour that repeats every week. Give it to an agent.

Free 30-minute call

Pick a time.

Frederik Theissen, founderTakes every first call himself.

Booking runs through Cal.com. Your name and email go to the calendar invite and nowhere else. Details in our privacy policy.

01Thirty minutes with the founder. We list the tasks that repeat: the inbox, the quotes, the follow-ups, the reports, the calls nobody picks up.
02A straight answer. Which of them an agent does reliably today, which not yet, and where to start. If nothing fits, we say so.
03A written next step. One fixed price for the first workflow. A person keeps the final say. What we build is yours.

What it looks like

One workflow first.
Built in your tools, with a person at the gate.

In plain terms

What AI workflow automation looks like in a small business

Every week the same hour comes back: the quote that has to be assembled from the price list, the inbox that needs sorting before anyone can start, the follow-up nobody sent, the report that gets copied together on Friday afternoon. That hour is where AI workflow automation earns its keep. An agent does the repeating part inside the tools you already run, and a person keeps the decision.

The word agent means a program that takes a task, works through it in steps, uses your systems along the way (mail, CRM, accounting, calendar) and writes the result back. You hand it work the way you hand work to a new colleague, with a brief and a check.

Workflows we build most often: incoming mail read, sorted and answered from the knowledge vault; quotes drafted from the price list and a short brief; supplier invoices read, checked and booked; missed calls answered and booked back; CRM records updated from call notes; weekly numbers pulled into a report a person edits. Each of these takes hours a week in a small firm.

Every step that touches a customer or money waits for a person at first. The office manager sees the draft quote before it goes out; the bookkeeper releases every invoice with a mismatch. As the log shows the agent doing the job correctly week after week, you loosen the gate, one step at a time. You decide the pace.

We start with one workflow, the one that repeats most and hurts most. It gets built in your tools, tested on your real data, and handed over with the code, the prompts and a runbook. When it has run quietly for a month, the next workflow follows. We work as an AI automation consultancy, so the plan comes from your week, and every system we build is yours.

Limits we name up front. A workflow with no fixed shape, where every case is a judgment call, is a poor first project; we pick something with a pattern. Agents need access to your systems, which means passwords, roles and a short conversation with your IT partner. And an agent that saves two hours a month is a bad investment; we say so on the call and suggest the inbox rule instead.

Questions we get asked

Which tools can agents work in?
Anything with an interface, an import or a browser: Microsoft 365 and Google Workspace, HubSpot, Pipedrive and most CRMs, DATEV, Xero, QuickBooks, Slack, Teams, your ERP. We check the specific stack before the quote.
How is this different from Zapier or Make?
Those tools move data between fields when a trigger fires. An agent reads the content, makes a judgment within the rules you set, drafts text and asks a person when unsure. Many of our builds use both: the plumbing from a tool like that, the reading and writing from an agent.
What does the first workflow cost?
A fixed price, quoted after the call and depending on the workflow and your systems. Every quote comes in writing with milestones. Founding-client terms exist for the first cohort.
Who maintains it?
You can. Everything is handed over with documentation, and your IT partner or an in-house person can run it. Most clients keep a monthly retainer with us for monitoring and changes.

Who you talk to

Frederik Theissen

A former academic researcher who has spent a decade building data and AI systems that run in production: analysis systems in neuroscience research, quantitative risk at a crypto hedge fund, and AI workflow automation for the analyst team of a Swiss market-intelligence firm. Building with large language models daily since December 2022.

Joinery builds the vault, the harness around it and the first agents that work from it, then hands it over as yours. I run my own company this way: one vault, a team of agents on top of it.

Founding-client terms available for the first cohort. Mention it on the call.

Start with the call.

Thirty minutes on the hour that repeats in your week, whatever it is. A clear answer, a concrete next step, and a working system within weeks if it fits.

Pick a time or write to hello@joinery-ai.com