Two Months, One Heat Pump Installer, and My Favorite Project This Year

TL;DR

• A small n8n automation request turned into the project I enjoyed most this year, in under two months.
• The lever wasn’t the technology, it was short communication paths, fast decisions, and a client who knew exactly what they wanted.
• Three lessons for anyone deciding between a big agency and someone like me for their next project.

The moment I realized this was going to be good

About two months ago, Damian reached out. The request was small: extend a few existing n8n automations for Wärme mit Konzept (WMK), add one or two new ones. Exactly the kind of project you knock out in an afternoon and forget about.

It didn’t stay small. That small request turned into a real software project: a custom order- and subcontractor-management tool with role-based access, automated PDF reports, and invoice processing. I wrote up the technical details and numbers, including the timeline and cost breakdown, as a case study on the TAISC blog. Here I want to write about something different: what made this collaboration one of the best projects I’ve had this year, and what I’m taking from it into how I want to keep working.

Lesson 1: Starting small is the best trust test there is

WMK could have opened with the big vision: “we need a complete order-management system.” Instead, a small, clearly scoped request came in first. In hindsight, that was exactly right, for both sides.

For WMK, it was a low-risk way to test whether the collaboration worked, without committing a large budget upfront. For me, it was a chance to show how I work before anyone had to make a five-figure decision. No pitch deck, no reference list, just a small piece of work done well.

If anyone asks me today how to best start with a new service provider: small, concrete, a task that’s done in one or two weeks. That tells you more about the collaboration than any kickoff call.

Lesson 2: Short paths beat big processes

What made this project stand out for me wasn’t the feature scope. It was how we worked. No middle layer, no project manager translating between me and the actual users. Direct alignment with Damian and his team, short feedback loops, decisions made in days instead of weeks.

That’s the part classic agency setups structurally struggle with: the more roles sit between the person who needs something and the person building it, the more information gets lost, and the slower every decision gets. At WMK, that chain didn’t exist. That’s also the core of what I described in Forward Deployed Engineering: working directly inside the client’s codebase and context, instead of passing requirements through multiple stations.

Lesson 3: The best project is one that isn’t finished

The sentence that stuck with me most from my last conversation with Damian: the collaboration isn’t ending, it’s just getting started. Next, we’re building on the existing system to cover further processes that previously ran through individual SaaS tools or spreadsheets.

That’s the real success metric for a project, to me, not the number of features shipped, but whether a client trusts you more after the first project than before, not less. In too many software projects I know secondhand, it’s the opposite: the first project ships, and then nobody reaches back out, because the experience was so exhausting nobody wants to repeat it.

What this means if you’re hiring for your next project

If you’re currently deciding who to bring in for a software project, here are three questions I’d ask in your position:

• Can you talk directly to the person writing the code? Or is there an account manager between you and the actual implementation who has to check internally first?
• How fast do you see a first working version after making a request? Days is a good sign, weeks is a warning sign.
• Does the first small project feel like a test for both sides, or like a contract that’s hard to get out of?

Those exact three things are what turned a small n8n request into a project I’d call one of my favorites this year. Not the technology, how we worked together.

If you’re at the point where spreadsheets and WhatsApp groups aren’t cutting it anymore: let’s talk, no obligation, ideally starting with one small, concrete task.