++AI automation and data engineering
An AI Automation & Data Engineer who connects your systems
The work that never gets done often sits between your systems rather than inside your product. Data that has to move from one tool to another, an integration that quietly stops working, a monthly report someone still assembles by hand.
This is the person who takes that work on and owns it: building integrations, making dataflows dependable and automating the steps your team keeps repeating, inside your team and at your pace.
You speak with Rens directly, usually within one working day.
This page covers integrations, dataflows and automation. See every AI role we provide
++Sound familiar
Four situations we run into often
Not every point has to apply. Two of them are usually reason enough to put someone on this.
- Someone copies data between two systems every week because there is no integration.
- Integrations exist, but nobody can say for sure they still ran today. Failures surface through a customer complaint.
- Monthly numbers are stitched together from exports, and nobody fully trusts the result.
- Scripts sit in production that one person wrote once, with no documentation and no owner.
This work rarely needs a separate project. It needs someone who picks it up structurally, next to your existing team.
++What this role picks up
Integrations, dataflows, reporting and recurring automation
Four kinds of work that almost always belong together in practice.
API integrations
Connecting systems through APIs and webhooks, with error handling, retries and logging, so a failure is visible before a customer notices it.
Dataflows
Moving data from source to destination and cleaning and checking it along the way, so what you rely on later actually holds up.
Reporting
Letting recurring reports come straight from the source instead of manual exports, with definitions that stay fixed.
Recurring automation
Replacing manual steps with a flow that does the same thing every day, with a clear owner and a place to see what happened.
AI comes in where it demonstrably solves something, for example interpreting text or unstructured input. Not as the starting point.
++What we work with
Python, SQL, APIs, webhooks, n8n, cloud and containers
Our public role and skill categories include AI, data and engineering profiles. These are the parts that keep coming back in them and that this role leans on.
Python and SQL
Python for custom integrations and processing, SQL for the work inside the data itself: modelling, checking and exposing it.
APIs and webhooks
Building and consuming REST integrations, handling webhooks, and dealing with authentication and rate limits properly.
n8n and similar tools
Building visible workflows between systems. Whether a workflow tool fits depends on the process, its risks and the systems already in place.
Cloud and containers
Running and shipping in the cloud with containers, plus logging and monitoring so you can see when something breaks.
We only name technology that appears in our public role and skill categories. What a specific profile brings is something we discuss per role.
++Choosing tools
n8n, custom code or a combination?
Start with the process, the risks and the technical setup you already have. n8n is not automatically faster, cheaper or better than custom code.
A workflow tool such as n8n
- A fit for
- Straightforward processes between common systems, visible workflows and relatively little custom logic.
- What to consider
- Error handling, version control, access permissions and who maintains the workflow.
Custom Python or JavaScript
- A fit for
- Complex logic, high volumes, unusual APIs, extensive testing or features that belong inside your own product.
- What to consider
- Maintenance, documentation, deployment and ownership.
A hybrid approach
- A fit for
- A workflow tool coordinates the steps while custom code handles the complex or critical logic.
- What to consider
- Clear boundaries, monitoring across both parts and one owner for the whole solution.
++Production readiness
From working demo to reliable automation
A workflow running once does not make it a dependable production solution. Check how it behaves with different inputs, failures and outages.
- Check the source data and who is allowed to access it.
- Test everyday cases as well as exceptions, including missing data and duplicate inputs.
- Handle errors and plan a safe fallback, such as stopping for human review.
- Add logging and monitoring so failures can be spotted and followed up.
- Store secrets such as API keys separately and grant only the permissions needed.
- Document the work, review the code and name a human owner.
++How we do this
Capacity inside your team, not a consultancy engagement
So you know upfront what you get and what you do not.
What you get
- A fixed developer working inside your team and your process
- Work in your repository, your environment and your way of shipping
- Someone who becomes the owner of the integrations and flows they build
- Direct contact with us about the collaboration
What it is not
- Not a consultancy engagement or an AI strategy document
- Not project outsourcing with an agreed scope and accountability for the result on our side
- Not a platform or licence you buy from us
- Not a few weeks of temporary help
You set the priorities and the technical direction. We handle local employment, the workplace in Prishtina and the local support around it.
++Lead time
Profiles within seven working days, contributing in two to four weeks
The steps are the same as for our other roles, with a profile shaped around this work.
01
A conversation about the work
We go through the systems you run, the work that happens by hand today and who is involved technically. That gives us the profile and the level.
02
Profiles within seven working days
You usually get matching profiles within seven working days. We check engineering skill, English and collaboration upfront.
03
You run the interviews
You speak with the candidates yourself, in whatever form suits you: a conversation, a technical session or a look at real work.
04
Start and onboarding
Once you choose, we arrange the contract, equipment and workplace. Within two to four weeks the developer picks up work from your backlog.
Lead time depends on the profile and on how quickly interviews get scheduled on your side.
++Pricing
The right profile. One clear monthly price.
From junior developer to tech lead. The monthly price follows the profile, the seniority and the technology you need.
Junior
Picks up tickets independently, with guidance.
from€2,495/mo
Medior
Builds features independently inside your architecture.
from€3,795/mo
Senior
Argues choices, guards quality, lifts others.
from€5,495/mo
Expert
Deep knowledge of a specific domain or stack.
from€6,795/mo
Tech Lead
Technical leadership and architecture for a team.
from€7,795/mo
Recruitment, local employment, payroll, the workplace in our Pristina office and local support are included. The final price depends on the profile, the experience, the technology and how scarce the role is. Any one-off costs are agreed upfront.
35 to 45% lower employer cost for medior and senior developers. Based on our representative 2026 employer scenario. See the calculation
++FAQ
Frequently asked questions
What kind of work does an AI Automation & Data Engineer pick up?
Is this an AI project or extra capacity?
When should you use n8n and when should you use custom Python or JavaScript?
How quickly can someone start?
What does this cost per month?
++Keep reading
Where to look next
Same way of working, different angle.
Tell me which work keeps getting stuck
Name the systems and the manual work that comes back every week. I will tell you which profile fits, what it costs per month and when this is not the right step.
No sales pitch. Just an honest view of what is realistic.
Rens GerritsenFounder · tre plus