AI
AI automation
It is rarely the big decisions that eat the time. It is the forty times a week someone copies something from one place to another and fixes the format on the way.
- Same-day reply
- Fixed price
- No lock-in
- Measured in time saved
- Starts in one team
- Runs in your systems
- Can be switched off
Service
AI automation
What I most often automate
- Extraction from documents: invoices, order confirmations, datasheets and contracts that arrive as PDFs.
- Categorising and routing incoming enquiries so they land with the right person immediately.
- Product data: descriptions, specifications and translations across many SKUs.
- Assembling reports currently built by hand from three systems every Monday.
- Cleaning up data typed in by humans over many years and never standardised.
When it pays off
The arithmetic is not hard. How many times a week does it happen, how long does each one take, and what does it cost to build. If payback lands within a year, it is usually worth doing.
There is a trap: tasks that feel annoying but only happen twice a month. They are rarely worth automating, even though everyone agrees they are stupid.
On errors
Automation with AI is not deterministic. That means there has to be a plan for the cases where it is unsure or wrong.
I usually build with a threshold: if the system is confident enough, it goes through. If not, the case lands in a queue a human looks at. That also gives an ongoing measure of how well it is really doing.
Deliverables
What you get
Concrete deliverables, not a report of recommendations someone else has to find time for.
- An automated workflow
Running in the systems people already use, not in a new tab someone has to remember to open.
- A threshold and a queue
If the system is confident enough, the case goes through. If not, it goes to a person. That also gives a continuous measure of quality.
- Before and after numbers
How long it took, how long it takes now, and how many cases still need a hand.
- An off switch
A way to switch it off without calling me. It sounds obvious and is the thing people most often ask for afterwards.
Process
How it works
Four steps. You know what happens when, and you can stop after any of them.
- 01
Find the task
We look for the work that gets done many times a week, follows a pattern and annoys the people doing it. That is where the first win is.
- 02
Measure the baseline
How long does it take today and how often does it happen. Without that number you cannot say afterwards whether anything improved, and it becomes one gut feeling against another.
- 03
Build and evaluate
I build a test set of real examples early and measure against it every time the prompt or model changes. It is the single thing most often missing in projects I take over.
- 04
Put it into production
Into the systems people already work in, with logging on everything and a way to switch it off. Plus an agreement on who owns it once I am gone.
Who does the work
I am the person you will be working with
My name is Mikkel and I have been building for the web since 2010. There is no project manager in between, and you will not get a junior on the job because the senior was busy. What you agree is what gets built, and I am the one building it.
Day to day I am a partner at Bigum&Co, and I have worked on more than 400 projects across ecommerce, pharma, construction and the public sector. Call or write — I usually reply the same day, and I will say straight away if the job is not a fit.
- Experience
- 16 yrs
- Projects
- 400+
- Rating
- 5.0
- Reply time
- Same day
In depth
AI automation
What is actually worth automating
The arithmetic is not complicated: how many times a week does it happen, how long does each one take, and what does it cost to build. If payback lands within a year, it is usually worth doing. If it is further out, there needs to be another reason — fewer mistakes, faster response to customers, or the task being so dull that it costs you in retention.
There is a trap in it: tasks that feel annoying but only happen twice a month. They loom large precisely because they are unpleasant, and they are rarely worth automating. Meanwhile there is nearly always a task nobody thinks about because it has become a habit — and it happens forty times a week.
So I always ask to observe for a week before we decide anything. Not a workshop where people describe what they spend time on, but the actual list of what got done. The two rarely look alike.
On errors and who is accountable
Automation with a language model is not deterministic. The same invoice can be read correctly nineteen times and wrongly the twentieth. That is not a bug to be fixed away — it is a property of the tool, and the answer is to build around it.
The way I do that is with a threshold. If the system is confident enough in its answer, the case goes through. If not, it lands in a queue for a person to look at. The threshold can be tightened or loosened as you see how it performs, and the queue doubles as a running measure of quality.
Accountability still sits with you. An automated process does not relieve the company of being in control of what it sends out. So the approval step stays on anything going to a customer or moving money, however well the system performs.
Reviews
What people say about me
I have never come across a developer as dedicated and capable as Mikkel.
Jonas Krogslund(translated from Danish) Mikkel is seriously good, helpful, lightning fast to reply — and on top of that he is a nice person.
Trine Handskemager(translated from Danish) Mikkel delivered competent, professional work setting up my website. Everything ran quickly and without a hitch.
Sisse Find(translated from Danish)
Services
Why me
- 16 years of experience
I have built for the web since 2010 and seen most of the ways a project can come off the rails. It makes the estimates more honest.
- 400+ projects
Across ecommerce, pharma, construction and the public sector. Chances are I have seen something close to your job before.
- You talk to the person coding
No project manager in between, and no junior on the job because the senior was busy. What you agree is what gets built.
Brands I have worked with
A selection. Some freelance, some through agencies, some across several years.
Questions
Frequently asked questions
- Do you use tools like Zapier or Make?
- For simple chains, yes, it is faster and cheaper. As soon as errors need proper handling or volume gets large, writing it as code becomes cheaper.
- Does this mean someone loses their job?
- In the projects I have done, the answer is usually no. What disappears is the part of the day nobody liked. But it is a decision you have to make, and I think it should be taken openly rather than sprung on people.
- How quickly does an effect show?
- On a bounded workflow there is usually something running within a month. The effect is measurable from day one, provided we measured the baseline beforehand.
- Can you automate something involving PDFs?
- Yes, and it is one of the task types with the most to gain. Invoices, order confirmations and datasheets as PDFs are still typed in by hand in a great many places.
- What if our systems have no API?
- Then it gets more expensive, but not necessarily impossible. Sometimes files can be exported and imported instead. I will say so if the detour gets long enough that it stops paying off.
- Can we adjust it ourselves afterwards?
- The parts about rules and phrasing, yes — I put those somewhere they can be changed without code. The integrations themselves need a developer.