I build production
AI systems
Agentic process automation, document intelligence, agent-built applications. No slideware — I build and maintain systems that run in production.
Why most AI projects fail.
Most companies that invest in AI end up with a tool nobody uses. Not because the technology is weak — it has never been better — but because the people who design the system don't do the work the system is meant to improve.
The person who closes your month-end in 90 minutes is that fast because she's done it two hundred times. She knows which line items move, which exceptions matter, and which vendor always sends the wrong invoice. None of that is written down. None of it survives a slide deck for the vendor. And none of it finds its way into a prompt on its own.
So the vendor builds something elegant. The employee quietly goes back to her spreadsheet. Six months later the system is shelfware and the budget is gone.
I build with the people who do the work. The system that ships is the one already running in their hands — minus the parts a machine does better.
Working AI systems, not slide decks
I don't consult from presentations. I build and maintain production AI systems — from process automation to entire applications.
Agentic process automation
I build automation systems on n8n with AI agents — and where low-code ends, I write custom microservices: APIs, scrapers, document converters.
Document intelligence & RAG
I process hundreds of thousands of documents — PDFs, scans, mixed formats: extraction, structuring, semantic search.
Agentic software development
Applications and backends built agentically (Claude Code + MCP) — always under the control of architecture, specifications and tests.
MCP, fine-tuning & architecture
I design MCP servers, fine-tune and select models for specific tasks, and take care of security and cost control.
Ready for AI in your company?
Book a free consultationWe're a fit if:
- 01
There's a team and a process owner — automation then scales the work of many people, not one.
- 02
The process runs weekly or daily, not once a quarter.
- 03
The work lives in documents, email, conversations, or across multiple systems.
- 04
The savings are worth at least one full-time role — or directly accelerate revenue.
- 05
The goal is a production deployment, not a free proof of concept.
If you're looking for an AI slide deck or a free PoC — I'm not the right address.
How we work together
From the first call to a working system — four stages, each with a clearly defined scope.
- 01
Free consultation
A 30-minute conversation about your processes. I point out where AI is likely to pay back fastest — no strings attached.
- 02
Workshop or process analysis
I work with the people who do the work every day. Together we pick the use cases with the highest return.
- 03
Implementation
I build iteratively, starting with the area that pays back fastest. First results show within a few weeks.
- 04
Maintenance & growth
I maintain and evolve the running system — and if you want, I train your internal automation team to build independently.
The scope and price of each stage are agreed before it starts — you know what you're paying for and what you'll get.
Implementations that work
Real projects: what I built, for whom, and the results.
Over 200 workflows on one platform — marketplace automation across 6 countries, agentic product research, custom microservices and an internal automation team trained to work independently.
Read the case studyA self-updating base of 538,000 tax interpretations, hybrid search, a research agent and opinion drafts. From ~12 hours of work to ~2 minutes.
Read the case studyFor a CFO-as-a-Service firm: an agentic content engine publishing one article a day, and a Coda-based finance operating system — a counterparty registry, an income-and-expense ledger fed straight from invoices, and a hub that shows who needs paying and how much.
Read the case studyThe tool that lets AI agents build working n8n automations — 22,600+ GitHub stars, 128,000+ users and 1.1M+ workflows built, including at Deutsche Telekom, PayPal, Mercado Libre, MIT and NYU. Built entirely agentically.
Read the case studyCost invoices arriving across seven team inboxes are found, read, de-duplicated and filed automatically — into a Google Drive archive and a Google Sheets register that's ready to hand to the accountant. Twelve connected n8n automations.
Read the case studyWhat clients say
I'm a certified n8n partner. These reviews come from the official n8n expert directory.
As a CEO, I've worked with plenty of technical vendors — Romuald is the one who moved the needle. He set up and maintains our n8n platform (queue mode, upgrades, rock-solid), built our Amazon scraping workflows and Supabase backend, then mentored me on using Claude Code to run the business — which genuinely changed how I work. He explains everything in plain language, so you're never in the dark. If you want automation that lasts and a partner who levels up your team, hire him.
Adrian G.CEO, mTeam Group We have been working with Romuald and AiAdvisors since January 2025 and continue to do so. He combines strong n8n, AI, integration and scraping skills with a practical, iterative approach. He turns complex ideas into simple first experiments, delivers reliably and actively transfers knowledge to our team. Thanks to this cooperation, automation and AI are becoming real internal capabilities at ESTYL.
Marcin DybickiCEO, ESTYLWorking with Romek has been a very smooth and professional experience. He has a strong understanding of business needs, proposes practical solutions, and explains complex technical topics clearly. Throughout the entire project, we could rely on excellent communication, timely delivery, and responsive support. I recommend Romek to anyone looking for a reliable and committed partner in automation and n8n-based solutions.
Łukasz KupryjańczykSenior Partner, Thedy & Partners
Romuald Członkowski
I build and ship production AI systems — from agentic process automation, through MCP servers and model fine-tuning, to document-analysis pipelines. I'm the creator of n8n-mcp — an open-source tool that lets AI agents build n8n automations on their own.
I help organisations move from AI experiments to architecturally coherent, maintainable production systems. I add the most value where a team and processes already exist — automation then scales the work of many people, not one person.
AiAdvisors- Creator and maintainer of n8n-mcp (22,600+ GitHub stars)
- Certified n8n partner
- World Bank expert advisor
- AiAdvisors founder with multiple production implementations
- Creator of "AiAdvisors | Romuald Członkowski" on YouTube
Common questions
Let's talk about AI in your business
Book a free strategic consultation and discover what AI can do for your company.
or write: romuald@aiadvisors.pl · +48 695 263 884