More · The fit

Agile delivery for teams already moving at AI speed.

The brief is not more ceremonies. It is a delivery system that catches up with an AI-assisted engineering team: clearer outcomes, autonomous teams, Product–Engineering alignment, and sustainable flow.

View the full CV See the delivery evidence
The short answer

This is a strong overlap.

I sit in the useful middle of this assignment: experienced enough in Agile transformation to see beyond the framework, technical enough to work credibly with engineers, and hands-on enough with Claude to understand why the old delivery assumptions are changing.

AI makes code creation faster. It does not automatically improve prioritisation, review, decision-making, stakeholder trust, or team health. Those become the new constraints. That is where I would focus.

15+
years delivering production software
4
years deep in agentic systems
115+
agents in the live AIOS workspace
23
AI-native sites in production
Requirement → evidence → move

Five reasons the work fits.

Each part of the brief maps to work I have done and a concrete way I would approach it.

01

Shape delivery for AI-assisted engineering

Help an engineering team turn dramatically higher coding throughput into reliable product outcomes.

Evidence

I run three live agentic platforms, a 115-agent workspace, and a 23-site AI-native fleet. Claude Code, skills, subagents, MCP, review gates, and versioned delivery are the operating environment—not a workshop topic.

The move

Redesign the flow around smaller changes, explicit context, fast review, automated evidence, and human accountability at the decisions that matter.

02

Drive transformation, not ceremonies

Evolve Agile practice across engineering without adding process for process’s sake.

Evidence

Fifteen-plus years of delivery includes digital transformation at Brittany Ferries, regulated AI delivery at Allianz, cross-region Agile delivery at RWS, and data-platform delivery at Sainsbury’s.

The move

Start with the system of work: map the real path from idea to production, find the constraint, and change only the policies and habits that improve flow.

03

Build autonomy with accountability

Coach teams to own outcomes and make good decisions without waiting for a facilitator.

Evidence

My delivery model uses a clear outcome, named ownership, acceptance evidence, visible decisions, and a human release gate. Agents accelerate the work; people remain responsible for the result.

The move

Coach through live work, clarify decision boundaries, and leave the team with an operating model it can run without the coach.

04

Join Product and Engineering

Improve collaboration and stakeholder management across the delivery boundary.

Evidence

I have worked as Technical PM, Delivery Manager, consultant, and hands-on engineer. That lets me translate business intent into executable slices—and explain technical constraints without hiding behind jargon.

The move

Create one shared outcome, one prioritised queue, explicit trade-offs, and short evidence loops with stakeholders.

05

Remove bottlenecks sustainably

Use AI throughput to improve the whole system rather than overload review, QA, or Product.

Evidence

The agentic delivery playbook I use pairs parallel execution with contract tests, WIP discipline, review roles, recoverable fallbacks, and small versioned releases.

The move

Measure flow end to end, reduce queue age and hand-offs, automate repeatable checks, and protect focus instead of rewarding permanent urgency.

Start with the work

The first six weeks.

No transformation theatre. Observe, change one constraint, prove the effect, and transfer ownership.

  1. 01
    Days 1–10

    See the real system

    Observe live delivery, map idea-to-production flow, interview Product and Engineering, sample the AI-assisted workflow, and establish a small baseline: cycle time, queue age, review load, defects, and blocked work.

  2. 02
    Weeks 3–4

    Change the constraint

    Co-design the smallest useful intervention—often slicing, ownership, review policy, stakeholder cadence, or the definition of evidence for AI-generated change. Trial it with one team before scaling.

  3. 03
    Weeks 5–6

    Embed and transfer

    Coach through live work, inspect the measures, adapt the model, and make team members the owners. Document the lightweight playbook so improvement continues without dependence on the coach.

A useful principle

AI moves the bottleneck. Coaching has to move with it.

When implementation accelerates, the constraint usually shifts to decision quality, context, review, integration, or stakeholder alignment. The answer is not to make stand-ups faster. It is to redesign the surrounding delivery system.

Flow
Lead time · cycle time · queue age

Is work moving faster, or merely starting faster?

Quality
Review churn · escaped defects · rollback rate

Is AI throughput producing trusted change?

Focus
WIP · blocked time · unplanned work

Is the operating rhythm sustainable?

Ownership
Decision latency · dependency age

Can the team move without waiting for permission?

Track record

Full CV

Enterprise delivery, transformation roles, product leadership, and the current agentic stack.

Operating model

Agent fleet

The live Discovery → Strategy → Design → Engineering → Distribution system.

Technical depth

Claude Code

The production workflow, skills, hooks, review patterns, and delivery guardrails.

Worth a conversation.

The strongest fit is the combination: delivery transformation, Product–Engineering translation, and current hands-on experience of how Claude changes the software system around the code.

Continue on LinkedInView Mat’s CV
Roll the dice