The team repeats the same manual workflow
Copy-paste between tools, re-typing data, manual document handling. I automate the specific process end to end so people stop doing it by hand.
Available for automation projects
AI automation
Private AI automation for small teams that want to keep their data
Your team burns hours on repetitive work: copying data between tools, drafting the same replies, searching scattered documents. I build focused AI automation that removes that busywork, and I design it so your data stays under your control instead of being handed to a third party.
A good fit for small teams and business owners who want practical automation with clear boundaries on where their data goes, not a generic chatbot bolted onto everything.
Where AI automation pays off
There is no internal automation at all
Everything runs on memory and spreadsheets. I build small, reliable tools that connect your systems and remove the obvious bottlenecks first.
You worry about where your data goes
Sending customer or business data to a public AI service is a real concern. I design setups where sensitive data stays on infrastructure you control.
Knowledge is scattered and hard to find
Answers live in old docs, tickets, and people's heads. I build a private search or assistant over your own content so the team finds answers fast.
Past AI experiments were unreliable
A demo worked once and then drifted or hallucinated. I add observability and guardrails so an automation is monitored, not just launched and forgotten.
What an automation project includes
A single, well-chosen workflow first
We start with one process that clearly costs time, prove it works, then expand. No boiling the ocean.
Private-by-design data handling
Sensitive data stays on infrastructure you control where the use case needs it, with a clear map of what goes where.
Integration with your tools
The automation connects to the systems you already use through their APIs, so it fits into how the team actually works.
Observability and guardrails
Logging, checks, and fallbacks so you can see what the automation did and trust it in production.
Documentation and handover
A written explanation of how it works and how to change it, so the tool is not a black box only I understand.
How an automation project runs
Each step is scoped and quoted before it starts. We prove value on a small piece before investing in a larger build.
Find the right workflow
We look at where time actually goes and pick one process with a clear payoff and acceptable data boundaries.
Prototype
I build a working prototype of that one workflow so you can see the result before committing to a full setup.
Production setup
We harden it: integrations, private data handling, observability, and guardrails, then put it into daily use.
Support and expand
I keep it healthy and, when it earns its place, automate the next workflow.
Automation pricing
Starting points for AI automation
Prototype
from €750 one-off
A working proof of concept for one workflow, so you can judge the value before investing further.
- One workflow, built end to end
- Runs against real (sample) data
- Clear recommendation on next steps
Private setup
from €4,000 per project
A production automation where sensitive data stays on infrastructure you control.
- Private-by-design data handling
- Integration with your existing tools
- Observability and guardrails
Custom tool
from €5,000 per project
A larger internal tool or assistant tailored to your process, delivered end to end.
- Scoped after the prototype
- Multiple workflows or data sources
- Documentation and handover
Support
from €750 per month
Keeping an automation reliable: monitoring, tuning, and small changes as your process evolves.
- Monitoring and health checks
- Tuning and small changes
- Cancel anytime
All prices are starting points in EUR and depend on scope and how much of the setup must stay on your own infrastructure. We always start small to prove value before a larger build. Working remotely from Spain with clients across Europe; other currencies quoted on request.
Relevant automation and platform work
Internal tools and product systems where automation removed manual work. Each links to a longer write-up.
INTERNAL PLATFORM
SPALAH CRM — Internal Business Platform
Full-Stack Engineer
Internal CRM and business management platform for a corporate training company — member management, course scheduling, and reporting.
View case study →- JavaScript
- Laravel
- MySQL
- PHP
- +3
SAAS PLATFORM
WProofreader SaaS — Subscription Management Platform
Lead Full-Stack Engineer
Full SaaS subscription platform built from scratch: Stripe billing via Laravel Cashier, OAuth sign-in with Jetstream team management, Sanctum API…
View case study →- AWS
- Docker
- Inertia.js
- Jetstream
- +12
Related reading
How I build and monitor AI agents and automation, from real projects.
Other things I can help with
WordPress & WooCommerce rescue
Your site is slow, updates break things, or customers cannot complete checkout. I find the real cause, fix the critical path, and leave you with a site your team can safely operate.
Laravel & SaaS development
Turn a product idea or an unfinished platform into software people can actually use and pay for — including subscriptions, onboarding, dashboards, integrations, and reliable releases.
Legacy modernization
Keep the business running while an old PHP, Laravel, or WordPress system is made safer and easier to change. Improvements are delivered in controlled steps, not as a risky big-bang rewrite.
Monthly technical support
For businesses that need someone to know the system and take care of it every month: updates, monitoring, fixes, and small improvements without briefing a new contractor each time.
Working with a business in Ireland?
I work with Irish clients remotely from Spain, with almost fully overlapping working hours. See the Ireland-specific pages for local market context and pricing.
Working with a business in Denmark?
I also work with Danish clients remotely from Spain. Spain and Denmark share the same timezone, so we overlap the entire working day. See the Denmark pages for local market context and pricing.
Frequently asked questions
What does "private" actually mean here?
It means we decide, per use case, where your data is allowed to go. Where the case needs it, sensitive data stays on infrastructure you control and is not sent to a public AI service. I am honest about the trade-offs: fully local models are less capable than the big hosted ones, so we choose deliberately rather than by default.
You work remotely from Spain. Does that matter for automation work?
No. The work is built through code, your APIs, and video calls, so it is remote by nature. I am in the CET timezone and overlap the European working day. What matters is that the automation is documented and observable, not where I sit.
Will I depend on you to keep it running?
No. The code lives in your repository, runs on your infrastructure, and comes with documentation. Support is optional. If you want to take it fully in-house later, you can.
Is this just a wrapper around ChatGPT?
Not by default. Some workflows are best served by a hosted model, some by a local one, and many need no large model at all, just solid automation. I pick the smallest thing that solves your problem reliably, and I tell you which parts use which.
How do you keep an AI automation from going wrong silently?
With observability and guardrails: logging what the automation did, checks on its output, and fallbacks when it is unsure. I have written about this for agent systems, and I apply the same discipline to client work.
Why start with a prototype instead of the full build?
Because AI automation is easy to over-promise and hard to judge from a pitch. A small prototype on one workflow shows the real value and the real limits before you spend on a larger setup.
Tell me what to automate
Describe the repetitive work eating your team's time and any data-privacy constraints. I read every message and reply personally, usually within one business day.