Open enrolment
Dates to be confirmed
We’ll find you the next one
Ask about datesOper8 — AI processes that learn from every decision.
An open method for redesigning a process from the ground up — or making an existing one faster with AI.
tap a node to read more
Outer loop · the customer teaches
Recommendation
What the AI puts in front of the customer at the moment of decision. Not a single answer, but an action — with its reasoning and the alternatives left open.
More in the methodCustomer response
The signal the customer sends back: accepted · corrected · ignored. Those three outcomes are what the flywheel learns from.
More in the methodSystem signal
An event from your systems — CRM, telemetry or billing — that enters the Decision Log automatically, with no one retyping or interpreting it.
More in the method
Core · Decision Log
Decision Log(Decision Log)
The record of each decision and its outcome. This is where both loops meet: the outer loop captures what happened with the customer, and the inner loop uses it to improve the next decision.
More in the method
Inner loop · the system learns
Quality gate(Quality Gate)
The tests every decision passes before it ships. Rolling back on degradation is an automatic step, not a meeting.
More in the methodMetrics(Metrics)
Six business metrics for an AI process, each paired with a counter-metric. The counter-metric catches the local maximum — the number goes up while the outcome gets worse.
More in the methodKnowledge base(Knowledge Base)
The rules, patterns and context the AI applies on the next turn. Not a static document — a snapshot of where the flywheel currently stands.
More in the method
Developed by Sergei Lipchanskii and Askhat Urazbaev at ScrumTrek, since 2008.
Where are you right now?
I don't know where AI would actually pay off
The team keeps talking about AI, but nobody can say where it would create value
We want to understand the method before hiring anyone
We need shared language to even have the conversation
How the Data Flywheel works, the autonomy spectrum and the five principles. 70 pages, free.
70 pages→ What we do
Request the bookWe want to learn the method ourselves
We have read the book and want to turn it into practice
We want to build the capability in-house and get the foundations right
We need a flywheel design and an eight-week plan
Two days with your team: choose a pilot, design the flywheel and leave with an eight-week roadmap.
2 days→ What we do
Book a workshopI don't know where to start
We have 15 to 200 processes — which ones go into a pilot?
Which one could pay for itself within eight weeks?
What does implementation actually look like?
You end up with a portfolio of initiatives, each with an ROI case and a roadmap.
1–2 weeks→ What we do
Request diagnosticsWe need help implementing it
We have no delivery team, or ours is already full
We want to bring the method in-house
We want to build a standing AI champions practice
We start the flywheel inside one of your processes, stay through the 8–12 week pilot, and hand the method over.
8–12 weeks→ What we do
Talk through your processI don't know where AI would actually pay off
The team keeps talking about AI, but nobody can say where it would create value
We want to understand the method before hiring anyone
We need shared language to even have the conversation
→ What we do
Request the bookWe want to learn the method ourselves
We have read the book and want to turn it into practice
We want to build the capability in-house and get the foundations right
We need a flywheel design and an eight-week plan
→ What we do
Book a workshopI don't know where to start
We have 15 to 200 processes — which ones go into a pilot?
Which one could pay for itself within eight weeks?
What does implementation actually look like?
→ What we do
Request diagnosticsWe need help implementing it
We have no delivery team, or ours is already full
We want to bring the method in-house
We want to build a standing AI champions practice
→ What we do
Talk through your processThe workshop — live the method for two days.
Day 1: we take your process apart, find where AI actually applies, and work out the economics of a pilot.
Day 2: we design the pilot, turn the flywheel through several cycles, and watch the agent get better in front of you.
By the evening: a Decision Log covering every iteration, and a defensible answer to whether the pilot is worth running.
In-house
Your process, your team
Scheduled around you
Talk through a workshopHow this works in practice
Haleon
Consumer healthBrief
from zero to working AI agents in two months, with flywheel gears at every stage.
Outcome
15 trained champions, the first agents in production, a Knowledge Base the team keeps filling, and a flywheel that turns without us.
✎Note: what was hard
the champions stalled for the first fortnight — not enough real work to practise on. We rebuilt the programme around the processes they were already running.
Read the case
US professional services firm
Public-sector bidsBrief
turn a decade of scattered proposal material into something the team can actually query.
Outcome
a Knowledge Base living inside the client's own environment became the team's primary workspace, with agent skills on top — including a weekly audit that keeps the base honest.
Read the case
- and 40+ more
Let’s
work together
We’ll look at your problem, pick the right format, and show where AI actually strengthens your operations.
50+
companies have gone through AI transformation with us — from diagnostics to working solutions.
© 2026 Kiberry — the European practice of ScrumTrek