EVAI CONSULTING Elbląg, Poland

Artem Kholomyanskiy

Business analyst on the client side: process audit, specification, implementation oversight. AI becomes useful only when the task is described precisely enough.

Artem Kholomyanskiy, AI consultant
1.0 // APPROACH

I start with process analysis, then design the solution and support implementation through to a stable result.

11+ years in enterprise IT in the energy sector, on the client side, as a business analyst and project manager. SAP IS-U, SCADA, metering systems, projects with high cost of error and strict quality requirements. Worked on company projects with a combined budget of over $10M. I use modern AI tools and automation platforms: Claude, Claude Code, n8n, API integrations, and other technologies suited to the task.

Philosophy Most problems in AI implementation start with a poorly described task and process.
2.0 // SERVICES

Services: from the first conversation to a working solution

All services →

3.0 // WHO FOR

Who it's for

Companies and teams that want to bring AI into real work processes.

A good fit for: small and medium business owners, department heads, companies preparing to implement AI, teams with a large amount of manual information work.

Ready to see where your department could save time? Briefly describe the task, and you'll get a reply within a day or two.

Write →
4.0 // PORTFOLIO

Practical examples of AI solutions

01

eSowa KPI Bot

Telegram bot for KPIs

02

Decor Engine

Catalog for e-commerce

03

LombardPRO

Pawnshop automation

04

Solo Founder OS

Virtual team on top of Claude Code

05

ELA City Bot

City assistant

Full portfolio →

5.0 // PROCESS

How an implementation project goes

For consulting and training the path is shorter: a free call, then sessions.

A description of the department and the task in two or three sentences. The reply says whether it is worth doing and where to start.
The focus is on real processes: what people do day to day and how long it takes. Result: a task map with an estimate of what can be automated. Audit formats and prices → · Automation payback calculator →
Solution architecture and a precise spec for each task: input, output, rules, what must never happen.
Rolled out one task at a time, accepted by whoever used to do it by hand, until the result is stable.
The solution is handed over with documentation. After that, as needed: one-off consultations or the next stage of automation.
6.0 // WHY I DO THIS

Where this interest comes from

I've always been interested in finding ways to make processes simpler and more convenient. Even before modern AI tools, I kept noticing repetitive actions, unnecessary steps, and awkward processes that could be changed. From everyday things to work systems: if an action repeats regularly, there's usually a way to make it simpler or more efficient.

The same happens in business. Employees perform many repetitive operations every day: searching for information, processing emails, filling out documents, preparing reports, transferring data between systems. I'm interested in finding the places where a process can be made simpler, faster, and more reliable.

My enterprise IT experience taught me to approach this systematically: first understand the task, then describe the process, and only then choose the technology. Today AI has become the next stage of this approach. It lets you build systems that take over routine work and free up time for decisions where a human is genuinely needed.

7.0 // PRODUCTS

What you can get right away

All products →

8.0 // FAQ

Frequently asked questions

Analyzes a company's actual processes, finds tasks that can be handed to AI or automation, designs and supports implementation through to a stable result.
Employees log their real tasks for a week. That data builds a process map showing where automation pays off and where the risk outweighs the benefit.
An open standard for describing tasks and requirements for AI systems, so AI behavior is predictable and clear to both humans and developers.
Claude, Claude Code, n8n, direct API integrations with models. The toolset is chosen to fit the task.
Express audit (one role): from 3,500 zł, 7–10 days. Department audit (up to 10 people): from 9,000 zł, 3–4 weeks. The final price depends on the number of people and processes and is fixed after a free 30-minute call. Prices are net, excluding VAT. Details →
A consultation answers a question: is automation worth it, what to check first, how to assess a vendor's proposal. The outcome is a recommendation and next steps. Training is for building the solution yourself: the participant works on their own task on their own computer, the consultant guides. The outcome is a working tool and the skill to build the next one. More about training →
9.0 // ARTICLES

Notes and articles

Notes on specifications, automation, and working with AI.

All articles →

Not sure what you need?

Describe what's happening in your department in a couple of sentences. You'll get an honest answer on whether it's worth doing and where to start.

By sending this form you share your name, email and message, which are needed to reply. Details: Privacy policy.

or write directly by email:

Write →

Book a 30-min call