Skip to content

Your AI problem is not a tools problem.
It's an engineering system problem.

I help CTOs and Engineering Managers find the organisational bottlenecks that AI makes louder, and fix the system underneath.

Book a 30-min diagnostic call See how I work

Free. No pitch. You'll leave with at least one named bottleneck.

Łukasz Ławicki, Engineering leadership advisory
  • Engineering Manager
  • Conference speaker
  • Co-host, Leadership Failed Successfully
  • Founder, Crusica

You bought the tools. The delivery didn't move.

  • More code, same throughput

    AI wrote twice as much code. Cycle time did not move an inch.

  • The bottleneck moved and nobody named it

    The constraint shifted out of coding and into review and problem definition, where nobody is measuring it.

  • Your metrics can't answer the CFO

    The dashboards you have cannot answer the one question you were asked: what did the AI budget return?

AI is an amplifier, not a fix.

If the codebase is weak, AI makes it worse, faster. If ownership is unclear, AI multiplies the confusion. The organisations winning with AI didn't buy better models. They fixed how they make decisions.

Systems over heroes. Bottlenecks over tools.

Every engagement runs the same four steps, in the same order. You always know which one we are in.

  1. Diagnose

    Week 1

    Interviews across the whole delivery chain, plus your own delivery data. No framework and no slides yet: first I need to see how work actually moves.

  2. Name the constraint

    Week 2

    One constraint, written in a sentence your CFO and your staff engineers both recognise. Everything else is noise until that one moves.

  3. Sequence the fix

    Week 3

    A 90-day plan ordered by what unblocks what, not by what is cheapest to start on Monday.

  4. Leave the system measurable

    Handover

    The metric set and the review cadence stay with your team, so the next constraint gets named without me in the room.

three ways to work with me

Who I work with CTO · VP Engineering · Engineering Manager. Product organisations of roughly 20 to 150 engineers.

  • Flagship CTOs and VPs of Engineering who are planning, scaling or rescuing an AI rollout

    AI Readiness Assessment

    You are investing in AI and cannot tell whether it is improving engineering. Often the constraint is not the AI.

    from €9,000

  • Engineering organisations of 20 to 100 where delivery is slower than it should be

    Engineering Effectiveness Review

    Delivery is slower than it should be and nobody can agree on why. A bounded diagnostic that finds where the time actually goes.

    from €7,500

  • Engineering leadership groups and the teams they are deciding for

    Workshops

    Two named days to choose from: where AI actually helps, or which constraint to solve first.

    from €3,000

See all engagements

where you can see the work

The consulting work itself is confidential. This is the part that is public: go and check it.

recent thinking

All insights

Not sure which problem you have? That's usually the first finding.

Book a 30-min diagnostic call

Free. No pitch. You'll leave with at least one named bottleneck.