How we got here
In early 2019, Fiona Greig and Alistair Drummond were both working as data engineers at separate Edinburgh firms. They kept running into the same problem: businesses wanted to use machine learning, but nobody on their teams could take a model from a Jupyter notebook to a live system that staff actually used.
They left their jobs that summer and set up a two-person consultancy from a shared desk in a Leith co-working space. The first client was a legal publisher that needed to classify thousands of case summaries by topic. That classifier went live in September 2019 and still runs today, processing roughly 400 new documents each week with 96% accuracy.
Word spread. A veterinary supply chain asked for demand forecasting. A Glasgow housing association needed automated document extraction for tenancy applications. By the end of 2020 the team had grown to five.
Today we are twelve people: six ML engineers, two data engineers, a project manager, a UX designer, a commercial lead and an operations coordinator. We work from an office in Larkin-le-Brekke, Scotland, and travel to client sites when hands-on collaboration is needed.
What guides our work
We wrote these down in 2020 after a project went sideways because we had not been honest enough about a data-quality issue. They have stuck.
Say it early
If the data is not good enough, or if a model will not solve the real problem, we tell you in week one. Silence costs more than an awkward conversation.
Ship, then refine
A model running in production with 85% accuracy teaches you more than a prototype stuck at 92% in a notebook. We get working software in front of real users as fast as we responsibly can, then improve it with live feedback.
Leave the keys
When a project ends, your team should be able to operate, retrain and troubleshoot the system without us. We write documentation, record walkthroughs and train your staff because our goal is independence, not dependency.
Measure what matters
Accuracy is easy to game. We agree on the business metric that actually matters to you (hours saved, error rate, revenue per recommendation) and report against that number every month.
The team
Twelve people, mostly based in Scotland. Here are the four you are most likely to work with directly.
Fiona Greig
Co-founder and technical lead. Eight years in NLP before starting the company. Leads model architecture decisions and client discovery sprints.
Alistair Drummond
Co-founder and commercial lead. Handles client relationships, scoping and contracts. Previously a data engineer at a Scottish energy firm for six years.
Priya Kaur
Senior ML engineer. Specialises in computer vision and document extraction. Joined in 2020 after completing her PhD at the University of Edinburgh.
Callum Firth
Project manager. Keeps timelines honest and makes sure clients hear from us before they have to chase. Certified Scrum Master with a background in software delivery for financial services.
Key milestones
June 2019
Fiona and Alistair register the company and take on their first client, a legal publisher needing NLP classification.
March 2020
Team grows to five. We deliver our first computer-vision project: reading handwritten veterinary prescriptions for a supply-chain client.
November 2021
Thirty projects completed. We move into a permanent office at 8 Kovacek Heights, Larkin-le-Brekke.
April 2023
We launch the Enterprise retainer model after three clients ask for ongoing embedded support rather than one-off projects.
January 2025
Sixty-two projects delivered. Team reaches twelve. Average client NPS over the past year: 71.
Where to find us
We are happy to meet in person at our office or visit your premises anywhere on the UK mainland. For an initial conversation, email or phone works best.
Address: 8 Kovacek Heights, Larkin-le-Brekke, Scotland, PI3 3KB, United Kingdom
Phone: +44 337 386 9050
Email: [email protected]