A team that has spent 18 years rebuilding how companies work. Now around AI.

Kiberry is the European practice of ScrumTrek. We packed three years of AI projects into the Oper8 method, and now run it through a book, a workshop, a diagnostic and full implementation. The people on your project are the people who wrote the method.

Get the Oper8 book →A 30-minute call about your process →

From Agile transformations to rebuilding around AI.

The Oper8 method is 18 years of watching how companies actually rebuild themselves, written down.

Since 2008, as ScrumTrek, this team has been helping companies rebuild how they build products. Those were the years of Agile transformations — when executives looked suspiciously at “giving up the plan”, and five to ten years later said “we have always worked this way”.

In 2023 the story repeated itself, this time with AI. The AI practice started inside ScrumTrek, and three years of projects ran into patterns nobody had written down: pilots that never reached production; a Rulebook nobody opened; a Data Flywheel that had stopped turning three months after launch.

In 2026 we packed that experience up and named it Oper8. We published the book and built the products around it: a two-day workshop, a diagnostic along the flywheel, and implementation as either a pilot or a full rebuild.

Sergei Lipchanskii, Askhat Urazbaev.

The strategic calls on a project are made by the authors of the method, not by a consultant they have delegated the work to.

Sergei and Askhat share an occupational deformity: the moment something in the industry cracks and everyone sighs, these two find the next trench and run at it.

Fifteen years ago they were explaining to executives that the familiar way of running projects had expired, and that teams needed more room and fewer sign-offs. The executives looked at them with suspicion. Then the same executives rebuilt their companies and acted as if they had always thought so.

Now the story is repeating itself, with processes and AI in place of management. So far the reaction is roughly the same.

Sergei Lipchanskii

Sergei Lipchanskii

Founder, Kiberry

Has spent twenty years breaking the way companies are used to working, and is still surprised they keep inviting him back. Over 20 years in IT, from writing code to introducing agile ways of working (Agile since 2010). Ran engineering teams his whole career.

Now works with AI agents every day: AI-first CRM systems, corporate sites, internal tools. He does not write the code — he runs the process and builds the harness. He has taught dozens of managers to do the same.

For businesses: identifying where AI can actually pay off, then choosing and rolling out the right platforms. For teams: hands-on training and longer-running schools. For executives: one-to-one mentoring on AI strategy.

Message Sergei on Telegram →
Askhat Urazbaev

Askhat Urazbaev

Co-founder, Kiberry

Graduated from MIPT with honours. For more than 15 years he has been making organisations work better through agile practice, digital transformation and product thinking, with Microsoft and with large banks and technology companies. Author of Lean Data Science.

When rebuilding management stopped being interesting, he moved on to rebuilding whole operating models around AI. The trenches got bigger; the habit of running at them stayed.

Has been advising companies on applying GenAI since 2023. On projects he carries the strategic decisions and the work with the executive team.

About the book Lean Data Science →Message Askhat on Telegram →

Who does the work on a project.

Behind the two founders is a bench of trainers, agent architects and DevOps engineers. A project draws three or four of them, picked for the format rather than for who is free.

  1. 01 · Role

    Lead

    One of the founders. Carries the strategic decisions and the conversation with the executive sponsor.

  2. 02 · Role

    Agent architect

    The technical design of the agent solution: which pattern fits, how it integrates with your systems, how the data is kept safe.

  3. 03 · Role

    DevOps engineer

    Deployment on your side: on your own infrastructure, or air-gapped where that is the requirement.

  4. 04 · Role

    Trainer

    The learning track: workshops for teams, programmes for process owners, support while the new process runs alongside the old one.

The size follows the format: a workshop is run by two people, a diagnostic by two, a pilot by four, a full rebuild by five in its busiest week.

Four principles we run projects by.

These four are what separates our projects from conventional AI consulting, and they are worth knowing before you start.

We start only when there is a named process owner on your side.

A Data Flywheel is not started by a consultant. It is started by the person who answers for that process. If nobody holds it at the start, we will say so and offer the workshop instead of a project: two days to work out who the owner is and to put the architecture together yourselves.

Anything touching your data stays under your control.

On your own infrastructure or in a cloud you control, inside the EU, and air-gapped where that is the requirement. Personal data goes only to models running inside your perimeter. Your own audit log, your own access policy, GDPR obligations written into the engagement rather than assumed.

The Oper8 book is free to read. Applying it is a separate piece of work.

We want the method used widely, and we are not holding anything back in the book. But between reading it and having a flywheel turning inside a real process there is a gap. That gap is closed either by your own work after the workshop, or by a project with us.

The workshop, the diagnostic and implementation are fixed formats with fixed scope and dates.

Each format has a fixed price and a fixed timeline, agreed before the work starts. What you get at the end is written down as an obligation, not as an intention.

How to reach us.

Get the Oper8 book →A 30-minute call about your process →

What happens after the call

Within one working day, a 30-minute conversation with Sergei or Askhat. We work out which step you are on — the method still on paper, choosing the process, or a process ready to be rebuilt — and point you at the format that fits: the book, the workshop, the diagnostic or implementation. If none of them fits, we will say so plainly instead of talking around it.

Related reading

The Oper8 method

The Data Flywheel, five elements of the flywheel, five principles — the method, in the open.

Read the method

The Oper8 workshop

A two-day intensive for teams: take a process apart and put an agent architecture together.

About the workshop

AI potential diagnostics

Self-assessment online, or a full process audit with a maturity map.

Book a diagnostic

Let’s
work together

We’ll look at your problem, pick the right format, and show where AI actually strengthens your operations.

Book a consultationTelegram

50+

companies have gone through AI transformation with us — from diagnostics to working solutions.

Kiberry

© 2026 Kiberry — the European practice of ScrumTrek

Privacy policy