About

What we do

We work in one area and take each implementation from start to finish: we connect AI to the systems and documents a company already has, and to the way its development team works. It takes three forms:

AI in business processes — automating one process, with sign-off at the end.
Connecting an AI assistant to your system — including through MCP, an open standard that lets an AI assistant use your system’s functions.
AI agents in your development team — AI agents are software that carries out the steps of a task itself; here they help the team write code.

Each project has one clearly defined scope: one process, a connection to one system, or the way of working within one codebase. At the start we agree which data and functions the AI can access and what should come out. We also agree who approves the result before anything is saved or sent.

Before building, we agree the acceptance criteria the result has to meet. We finish by handing the system over to the person who takes it on — with the code or configuration, documentation, tests and a maintenance guide. Practice with the system is part of every implementation, so that no dependency on us is left behind when the project ends — we do not run training or workshops on their own.

AI implementation — from checking whether an implementation is worth doing to acceptance and handover: AI implementation.
How we work — phases, decision points, and the moment you can walk away: delivery methodology.
What we don't do — the conditions under which we advise against a project: the refusals.
Data and compliance — where data goes and who is answerable for what: security and GDPR.

Who founded ploch.ai?

The company was founded by Krzysztof (Kris) Ploch — a software architect, developer, and technology enthusiast with over 20 years’ experience in software design and development, business process automation, systems integration, team management, and product delivery.

Kris started using AI before the technology became widespread — it was helping him in his work many years ago. At his last company, he was one of the people who introduced AI agents as the core way of working; within a year, more than 70% of the code in the team he worked in was produced with their involvement. That is the result of one team in one organisation, not a benchmark for other businesses or a promise of the same result in your company. It is also the experience that led him to make AI implementation his profession. Before that, he spent years working on software systems used day to day by companies across a wide range of sectors and of all sizes — from small and medium-sized enterprises employing several dozen people to global corporations with over 20,000 staff.

That is why we treat process automation in an operating business first and foremost as connecting the solution to the company’s systems and maintaining it, and only then as a question of choosing the AI model (the engine behind tools such as ChatGPT) or writing code.

Kris also maintains public repositories on GitHub, where he shares the code and ideas he creates in his spare time.

Registration details

Registered in Poland and operated from Lublin and Warsaw; we work with companies across the European Union. Full contact details including the address are on the contact page.

Legal name: Actum Krzysztof Ploch
NIP (tax ID): 5222208821 (EU VAT: PL5222208821)
REGON (statistical number): 015269213
Operating since: June 2024

Frequently asked questions

Who will actually work on our implementation?

ploch.ai runs every implementation from start to finish — from the first conversation, through checking the scope, access and test materials, to acceptance and handover. You therefore work with the company that designs and builds the solution, not with an intermediary who passes the work on. When a project needs additional specialist skills, we bring them in to match its scope.

How long have you been operating?

The business was registered in June 2024. The engineering experience these implementations rest on is older than that — it comes from working on software systems that businesses use day to day rather than from courses about AI.

What do you do, and what do you not do?

We automate one process at a time, connect AI assistants to company systems (including through MCP, an open standard that lets an AI assistant use a system’s functions), and set up AI agents in development teams — from the viability check, through building and testing, to acceptance, handover and practice with the system for the team that takes it on. We do not build models from scratch, do not do marketing based on bulk outreach, and do not take projects where an honestly calculated saving will not cover the cost of keeping the system running. That second list is written up separately, along with the conditions under which we advise against a project.

Are you an agency or a software house?

We are an AI implementation company. We start with one business process and check whether an existing tool can meet the need. Before building anything, we agree the scope and the acceptance criteria. If a custom solution does not pay off, we recommend a simpler option or advise against the project.