Skip to content
azimut

Domain Lab · Adaptation

Learning Architecture LabA program for reflection, experiments and lessons learned

Learning Architecture Lab is our program for managers and team leads on how a team learns from its own experience. Over 6–8 weeks, a cohort of 8–12 people works through three two-and-a-half-hour labs, applying each one with their own team between sessions.

Duration
6–8 weeks
Rhythm
one 2.5-hour lab every 1–2 weeks
Cohort
8–12 participants (15 at most)
Format
on-site or online

Who it’s for

When does a team need Learning Architecture Lab?

The problem usually arrives as one sentence: “it happened again, for the third time.” Day to day, it shows up like this:

  • The same problem keeps coming back for months, with slightly different people involved.
  • After a project, the conversation about what went wrong quickly turns to who is to blame.
  • Lessons learned end up in a document nobody opens again.
  • Improvement ideas get discussed for months and never become a test.
  • A pilot ends with “some people liked it, some didn’t,” and nothing gets decided.
  • One team has found a better way, and the other teams running the same process don’t know about it.

Is the problem more that people don’t say what they see? Dialogue Lab →

Want all four domains, for managers from several teams? Manager Operating System →

The journey

What do participants work on in the three labs?

Learning here doesn’t mean courses. It means a team’s ability to adapt: examining its own experience, testing small changes and updating its routines. Experience piles up regardless; what is usually missing is the mechanism that turns it into better practice. The three labs work in order: reflection, then experimentation, then integration. Each one starts from participants’ real situations and ends with something to apply with their team the following week.

  1. Lab 1

    From experience to lesson

    Reflection & Sensemaking

    A four-part review of a real event: the facts, what we assumed, the causes in the process, and what we do differently. The difference between a lesson and an opinion, and a short review facilitated even when the team wants to “just move on.” Each participant leaves with their first review done on their own case.

  2. Lab 2

    Test, don’t debate

    Experimentation & Adaptive Action

    An improvement idea turned into an experiment: one change, a fixed timeframe, a hypothesis, and the evidence agreed before the test starts. The decision at the end (keep, adapt or stop) made on evidence rather than impressions. Each participant leaves with a real experiment and its review in the calendar.

  3. Lab 3

    The lesson becomes routine

    Knowledge Integration & Team Learning

    A 20-minute after-action review, and a lesson’s path into routine: where it lives, who integrates it, and how we check it changed practice. The lesson taken to the other teams that need it too. Each participant leaves with a real lesson on its way into routine.

The order is deliberate: a good experiment starts from a clear lesson, and a lesson only goes into routine once it has been tested. The first lab lays the ground for the other two.

Between sessions

What happens between the labs?

The lab is the smaller part. The change happens in the weeks between sessions, in the real work.

Before each lab
a short microlearning: one video and a one-page job aid.
AI practice, twice a week
10–15 minute sessions with an AI coach on real situations: a guided review of a recent event, an experiment refined before launch, a stakeholder who wants certainty before any change. The feedback is built on the models from the lab.
A field assignment
every lab ends with actions for the following week: a 15-minute review of a real event, an experiment launched with its evidence agreed, a lesson taken all the way into routine. The next lab starts from what happened.
Real cases
before each lab, the sponsor sends us a live situation from the organization: a problem that has kept coming back for months, a process everyone knows is broken, an expensive lesson that never made it into any routine. We work through it together in the lab.
The participants’ own managers
get a briefing at the start and, every week, one question to use in conversations they are already having, such as “What’s the smallest test that would settle this?” when an improvement idea comes up. Their job is to reinforce, not to assess.

How it runs

The same spine, adapted to the program.

DiagnosticLive LabsAI PracticeWorkplace ApplicationManager ReinforcementReview

The program spine: from diagnostic to review

Measurement

How is progress measured?

Baseline
a questionnaire on the three capabilities and on how the organization learns, completed before the first lab.
Along the way
after each field action, a two-minute reflection log shows what is being applied and where it gets stuck.
At the end
the same questionnaire, two weeks after the last lab, in which participants also rate how much they feel they have changed on each capability. The line manager observes the same behaviors twice: before the first lab and at the end.
For the sponsor
a cohort report showing each capability before and after and the patterns that keep recurring, discussed in an executive review. How much participants feel they have changed is reported next to the measured change and the managers’ view, and never goes into the score.

The questions are the same before and after, so progress can be compared rather than described. The report also counts what can be counted: how many structured reviews were held, how many experiments had their evidence agreed in advance, and how many routines were updated.

Frequently asked questions

Three 2.5-hour labs, one or two weeks apart, plus two weeks of application after the last one: 6–8 weeks in all. Between labs: a short microlearning before each session, AI practice twice a week for 10–15 minutes, and field assignments with their own team.

Managers, team leads and functional leaders, in a cohort of 8–12 people (15 at most). It fits when the same symptoms show up across several teams: problems that keep repeating, reviews that hunt for culprits, changes discussed but never tested, lessons that don’t travel. If you want all four domains for managers, Manager Operating System is the right format.

A course delivers content over a day, and a lessons-learned session usually produces a list. Learning Architecture Lab builds behavior over time: every lab is followed by application with the team, practice, and questions from the line manager, and progress is measured before and after with the same questionnaire. And it doesn’t stop at the lesson: it takes it into the team’s routine, with someone who integrates it and a date to check.

Yes. One lab, with practice and a guide for managers, is the simplest way to test the approach; two labs cover the most pressing capabilities. Measurement grows with the format: for a single lab, a short pulse before it and again two weeks after, with the same questions; for two labs, a short before-and-after test on the capabilities covered; for the whole domain, the full questionnaire. From two labs up, the line manager also observes the same behaviors at the start and at the end.

A small, precise role of about 15 minutes a week. They get a briefing at the start and one question a week to use in conversations they are already having with their people. They reinforce rather than assess. At the start of the domain and again at the end, they observe the same behaviors, and their concrete examples go into the sponsor report without names.

AI is the practice partner between labs; it does not replace facilitation. Each participant rehearses their own real situations with a coach set up on the program’s models. The conversations stay private: we only see the reflection log, and the sponsor gets cohort-level results, not individual assessments. We agree the tools and the rules of use with you before the start.

Yes. Labs run at your offices or online, in Romanian or English. Online, the drills happen in small groups and the live case is worked through together on a shared whiteboard.

Pricing depends on the format (one lab, two, or the whole domain), the number of participants, the language, and how much we tailor the cases. We give you a concrete price after the first conversation, once we know what you want to change.

Further reading

Start with a conversation

Tell us the behavior you need to change. We’ll show you the system that changes it.

A 30–45 minute diagnostic conversation, no pitch deck. You’ll leave with a sharper read on the capability gap, whether or not you work with us.