About our Artificial Intelligence consultancy

How we got here

Trusted Ai Wave started in 2018 when our founder, Rowan Fairley, left a data-science role at a large Edinburgh insurer. The trigger was a familiar frustration: months of model development followed by a deployment that never happened because no one had planned the engineering side. Good models were dying in Jupyter notebooks.

Rowan hired an ML engineer, then a second. The three of them took on small projects with a simple promise: every model we build will run in production within eight weeks, or you do not pay for the modelling phase. That constraint forced us to pick problems carefully, audit data before writing any code, and package models as deployable artefacts from day one.

Seven years later the team has grown to nine people. The promise has not changed. We still turn down projects where the data is not ready, and we still quote fixed prices. Forty-seven models are running in production for nineteen different companies. Fourteen of those models have been live for over three years without a single retraining cycle, which says more about our data-validation process than about any clever architecture.

What we care about

These are not aspirational slogans. They are rules we apply to every project decision.

• Fixed-price honesty

  We quote a total cost before work begins. If scope changes, we re-quote and you approve the new number. No surprise invoices, no hourly overruns.

• Data first

  We audit your data before proposing a model. If the data cannot support the accuracy target, we say so and suggest what to collect instead.

• Production or nothing

  A model that lives in a notebook is a research project, not a product. We deploy every model as a containerised service with monitoring and alerting from day one.

• Knowledge transfer

  When the project ends, your team gets runbooks, feature registries and documented pipelines. You should be able to retrain the model without calling us.

The team

Nine people, all based in Scotland. No subcontractors, no offshore hand-offs.

Rowan Fairley

Founder and lead data scientist

Former senior data scientist at Scottish Widows. Twelve years of experience in predictive modelling for insurance and retail. Holds an MSc in machine learning from the University of Edinburgh.

Priya Anand

ML engineer

Specialises in model deployment, CI/CD pipelines and cloud infrastructure. Before joining us she built real-time fraud-detection systems at a London fintech. She writes most of our Docker and Kubernetes configurations.

Callum Rennie

Data engineer

Responsible for data pipelines, ETL jobs and feature stores. Callum joined straight from a computer-science degree at the University of Glasgow and has been with us for four years. He maintains the automated data-quality checks that run at the start of every project.

The rest of the team includes two additional ML engineers, a front-end developer, a project manager, a QA analyst and a part-time finance administrator. On larger projects we pair engineers so that code review happens in real time rather than after the fact.

Where to find us