Selected engagements & outcomes

The work behind the numbers.

A selection of operating challenges, the decisions behind the work and the results. Client sensitive details have been intentionally kept out.

01
Sama / Quality leadership

Building a quality function that could support enterprise scale.

As work expanded across projects, quality needed a clear structure and a consistent way for teams to learn from errors, prevent rework and improve output.

The challenge

Managing quality across many project types and large teams creates variation. A useful quality function must be connected to production decisions and client expectations, rather than operate only as a downstream audit.

The approach

  • Built a dedicated management and analyst structure.
  • Improved task flows and review practices.
  • Linked recurring errors to coaching and process changes.
  • Established clearer ownership for quality performance.
13Quality managers
200+Quality analysts
20% → 5%Rework reduction
02
CloudFactory / Client delivery

Growing the delivery team while improving quality.

Rapid team expansion can put quality, onboarding and supervision under pressure. In this engagement, delivery capacity and output quality improved in the same period.

The challenge

Scaling a client operation required more than recruitment. Quality expectations, performance visibility and team support needed to hold as delivery volume increased.

The approach

  • Coordinated the expansion of the delivery workforce.
  • Strengthened quality monitoring and feedback.
  • Reinforced performance expectations as the team grew.
  • Kept attention on client outcomes during expansion.
30 → 300+Team size
60% → 98%Output quality
3 monthsImprovement period
03
Sama / Enterprise programme

Keeping a multi geography programme accountable.

Large programmes need a common view of performance, a reliable operating cadence and timely decisions across distributed teams.

The challenge

Teams operating across locations can develop inconsistent views of productivity, customer requirements and priorities. Coordination becomes harder when new queues or requirements are introduced.

The approach

  • Managed programme execution across three geographies.
  • Coordinated workforce, delivery and quality stakeholders.
  • Used routine performance reviews to identify delivery risks.
  • Worked across technical and operations teams on process changes.
2,500+Team members in programme
3Geographies
04
Recent work / AI data operations

Making specialist AI data review easier to run.

Computer vision and document AI workflows produce difficult review decisions that cannot be solved through a single accuracy score.

The challenge

Video, LiDAR and OCR datasets each introduce different failure modes. Reviewers need calibrated definitions, clear escalation routes and processes that distinguish a systemic problem from an isolated error.

The approach

  • Designed reviewer onboarding and assessment processes.
  • Introduced calibration and consistent error categories.
  • Managed throughput and quality expectations together.
  • Coordinated operational feedback with client requirements.

Specific client information and operational metrics are omitted for confidentiality.

What connects the examples

Scale requires a system that learns.

The common thread is building an operation that can see its own performance. That means agreed standards, usable data, a way to address failure and clear accountability for improvement.

A dashboard cannot repair weak handoffs. A larger workforce cannot replace calibration. And a good pilot does not guarantee repeatability. The operating model has to keep pace with the work.

Read my thinking