Automation, AI agents, software and websites for small and mid-sized businesses in Hamburg, at fixed prices and with one contact who stays the same from the first call to the running system. Kasoria sits a few minutes south of the city limits.
The first call is free and runs over video. Afterwards you know whether the effort pays off for you.
Most people searching for an AI agency in Hamburg are not really searching for AI. They are looking for an answer to a question that came first: why has someone in the office been retyping the same numbers from an email into another program for three years, and why is that the person who never has time.
This page gives you the answers that come before an enquiry: what AI actually means in the daily work of a Hamburg business, which of the three shapes fits which problem and where your data stays along the way. How you work out in advance whether a process is worth it is in here too. On top of that come the processes businesses here actually ask about, and four projects from the region you can look at.
Behind Kasoria is one person rather than an agency with floors. You talk to the same developer who also builds your automation, from the first call to the running system, and that is why the assessment you get has not passed through a sales loop.
In conversation the word AI almost always means one of three very different things, and mixing them up gets expensive, because the three differ a lot in price, build time and risk.
The difference between the first two matters more than it sounds. An automation does the same thing on every run: an order arrives, it creates the record, writes the confirmation and attaches the invoice. When it does go wrong, you can point at the step. An agent decides again on every run. That is its advantage as soon as the input is messy, so with emails, call notes or photos of a delivery slip. It is its weakness as soon as the result has to be correct, because in case of doubt it will invent a plausible number instead of staying quiet.
Anything you can write down as clear if-then sentences belongs in an automation. Anything you can only explain with "you will see it when you look at it" is a case for an agent, and then with a person who signs off on the result. The longer version is in the article on AI agents and automation.
Most businesses need the first one now and the second one later. Which process comes first is a question of its own, and it matters more than the choice of tool, because the order in which you automate decides whether the project continues after step one or quietly stops.
The same calculation sits in front of every quote, and you can do it without any outside help. You need three numbers: how long the process takes once, how often it happens a month and what an hour costs in your business.
The numbers above are made up, the structure is not. Put your own in and you will see within 2 minutes whether the conversation is worth having at all.
Put in your three numbers. Everything is worked out in your browser, nothing is sent anywhere and nothing is stored.
That is the ceiling, not yet the result. Take the build off it and count in the ongoing running, and what is left is the real number.
One part is missing from that calculation, and quotes like to leave it out: an automation is not a piece of furniture, it is something that runs. Interfaces change, passwords expire, a vendor renames a field. Budget for running and support from the start, or month seven arrives with a process standing still that nobody is watching.
And the honest end of that calculation: below roughly two hours a month, almost nothing is worth building and then maintaining. Processes that small are better collected until three of them together make one project. The processes that carry Hamburg businesses past that threshold fastest are further down, sorted by industry.
This is the question that comes up most often in Hamburg businesses, and it comes up for good reason, because a language model normally means that text leaves your house. There are three ways to handle it, and they differ less in price than in effort.
Whichever route you take, there is one line I draw in every project: credentials, complete personnel records and anything covered by professional confidentiality do not go into somebody else's model. That is not caution, it is the condition for the whole thing holding up when it is questioned.
This part is missing from most agency pages, because it sells nothing. It does save you the three most common bad purchases.
Anyone telling you otherwise is selling you the project rather than the result. The most honest sentence on the subject is this one: a good automation makes a task smaller, and the task almost never disappears completely.
In an AI project, nearness means one thing above all, and that is knowing the processes. What comes up at a freight company on the Elbe, what a drywall contractor from Harburg still has to write in the evening and how a tender in the north gets checked is something you know from projects here rather than from a template. The businesses further down this page are exactly that.
On top of that comes the unglamorous part: the same time zone, the same language, a registered office a few minutes south of the city limits that you can look up in the commercial register, and a contact who is still the same person in the third call as in the first.
The work itself runs over video with a shared screen, and for a simple reason: the processes in question run inside your programs, and that is where you see them click by click, more closely than at a meeting table. It saves you the appointment in the calendar and the drive, and the attention your case gets stays the same.
The processes below come from real conversations with businesses in Hamburg and the surrounding area. They are not a list of what would be possible, they are what people here ask for.
Order emails with addresses, dates and weights get typed into the system by hand. An agent reads the email, an automation creates the record, a person confirms it. What that looks like for a courier service in the north is in the case study on SPEDO.
Measuring, quoting, following up, all of it in the evening at the kitchen table. Text blocks and prices become a draft you check instead of writing. A Hamburg example of the step before that is Prima-Max.
The same twenty questions every time, never in the same words. Sorting and drafting answers works well, deciding stays in the house.
Incoming documents get recognized, named and filed so that searching does not end in an inbox. This is where a model on your own server is usually the right route.
Hours, tasks and feedback live in four tools. An automation brings them together instead of somebody collecting screenshots every Monday.
Documentation duties eat the end of every shift. Whatever can be formalized gets prepared, checking stays with people, and the data stays in the house.
Four projects from Hamburg and the north. Three of them are written up as case studies, with the starting point, the approach and what came out at the end.
Drywall and interior construction, more than ten years in business and practically invisible online. Five trades that nobody could tell apart before now have a page each.
Read the case studyScrap and metal trading since 1948, and practically nothing came in over the old website. Today more than half of the callers name the site as the reason they got in touch.
“We ask our customers on the phone how they found us, more than half name our new website.”
Logistics and courier services, strong on referrals and regularly out of the running in formal tenders. Today a presence that holds up when a third party checks.
“Really happy with it! We finally come across as professional when we bid for contracts.”
Certified disposal business for electronic waste. Relaunched alongside its sister company, with a clear route to a commercial enquiry instead of a phone number in the imprint. How that relaunch went is written up in the Heinz Bollow GmbH case study.
Four steps, and after the second one you can walk away without a contract hanging over you.
Thirty to sixty minutes over video or phone. Afterwards you know whether your process is a case for an automation, for an agent or for neither.
On a shared screen we look at how the thing really runs today, click by click. What comes out is a proposal with a fixed price that says what gets built and what explicitly does not.
Built in small steps so you see early what is taking shape. Handover includes a briefing and documentation that somebody who was not there can follow.
Monitoring, changes and the repair when an interface moves. No minimum term, because a process that only keeps running because of a contract helps nobody.
Four cases where I advise against it or send you elsewhere. They cost me work, and they are the reason you can believe the rest.
If none of the four applies to you, the conversation is worth having. It costs nothing and often ends with a recommendation that is cheaper than a project.
The projects run over video with a shared screen, because the processes live in your programs and that is where you see them click by click. It saves you the appointment and the drive through the Elbe tunnel, and the attention your case gets is the same as at a meeting table.
The same as anywhere else, because the price depends on the process rather than on the postcode. Every service has three fixed tiers, and the figures are published on the service pages so you know the order of magnitude before we talk.
The processes people here ask about most often are further up, sorted by industry: logistics and port, trades and construction, retail, law firms, agencies and care. The industry matters less than the process, because an order email gets retyped by hand at a freight company just as it does at a wholesaler.
One person. That is a drawback if you need twenty people at once, and an advantage if you are tired of explaining your project again in every meeting. Whoever gets in touch talks to me from the first call to the running system.
That is your decision, and the three possible routes are in the section on data protection above. For sensitive material the model runs on a server you control. For everything else a vendor with a data processing agreement and a European region is usually enough.
In most cases yes, either through the program's interface or through whatever route it offers for exchanging data. Where a program is closed, I say so in the first conversation rather than discovering it mid-project.
A single automation is usually standing within one to two weeks, an agent connected to your systems within four to eight. The longest part is rarely the building, it is the question of how the process should actually run.
Yes, that is the second half of the work: websites and visibility in search engines and AI answers through SEO and GEO. For customers searching locally that is often the lever that comes before any automation.
One call, thirty to sixty minutes, over video or phone. At the end you know whether the process is worth it, roughly what it costs and whether you need anybody for it at all.
Free and without obligation. If I advise against it, I will tell you why.