Stellar AI Professionals started in 2018 when our founder, Roisin Gallagher, left a data-science role at a Belfast fintech firm to consult independently. Her first client was a local logistics company drowning in spreadsheets. She built a route-optimisation model in Python that saved them roughly £4,200 a month in fuel costs. Word spread.
By 2020, Roisin had brought on two more engineers. The three of them shared a co-working desk in Mueller-upon-Morissette and focused exclusively on tabular-data problems: churn prediction, pricing optimisation and anomaly detection. That narrow focus let them deliver faster than larger agencies because they were not context-switching between web design, SEO and AI.
Today the team has grown to seven. We still only do AI and machine learning. If a client needs a new website or a marketing campaign, we refer them to partners we trust. Staying in our lane means every project gets engineers who have solved similar problems before, not generalists learning on the job.
These are not aspirational slogans pinned to a wall. They are the criteria we use when making trade-off decisions on real projects.
We run a two-week discovery sprint before quoting any project. If the data does not support the outcome you want, we say so in writing and suggest alternatives. Roughly one in five enquiries ends at this stage because the honest answer is "not yet."
Slide decks do not reduce costs. Working code does. We write clear technical documentation, but only after the model is running in production. Our clients judge us on deployed results, not on the thickness of a strategy report.
We train at least one person on your team to retrain, monitor and troubleshoot the model we build. If we get hit by a bus, your system keeps running. Every engagement includes a hands-on workshop tailored to your engineers' skill level.
Your datasets never leave your infrastructure unless you explicitly request cloud training. We work inside your VPN or on-prem servers. Model weights, training logs and evaluation metrics belong to you, full stop.
Seven engineers, one operations manager. No sales team, no account executives. The person who scopes your project is the same person who writes the code.
Founder and lead engineer
Roisin spent five years building fraud-detection systems at a Belfast fintech before starting Stellar AI Professionals. She specialises in tabular data modelling and has a particular interest in explainability methods like SHAP and LIME.
Senior ML engineer
Declan joined in 2020 after a decade at a Dublin-based analytics consultancy. His focus areas are natural language processing and document extraction pipelines. He maintains our internal OCR benchmarking toolkit.
ML engineer
Amara holds an MSc in computational statistics from Queen's University Belfast. She focuses on time-series forecasting and has deployed demand-prediction models for three retail clients since joining the team in 2022.
The rest of the team: Conor (MLOps and infrastructure), Siobhán (computer vision), Priya (data engineering) and James (operations and client coordination). Between us we have shipped over 40 production models across retail, insurance, healthcare and logistics.
Tell us about the problem you want to solve. We will schedule a 30-minute call to discuss whether AI is the right approach and, if so, what a realistic timeline looks like.
Contact us