Summary
Mattias Arro is an AI & ML architect with over two decades of software experience and more than a decade focused on machine learning, currently designing production-ready AI systems and training programs at Pragmatiq AI. He has led MLOps and productionisation efforts across startups and enterprises, building platforms on Kubernetes, Polyaxon, Kubeflow and major cloud providers to remove manual steps while preserving data scientists’ exploratory workflows. As a founder/CEO he shipped autonomous web-extraction AI and ran an ML engineering consultancy, combining product-minded engineering with hands-on model development in OCR, document understanding, transformers, VLMs and multi-modal pipelines. His background spans end-to-end ML—data pipelines, multi-objective training, retraining workflows and deployment—across GCP, AWS, Azure and databricks ecosystems. Educated in data science at KTH and Eindhoven with a BSc in Computer Science and a BA in Semiotics, he brings both technical rigor and a nuanced appreciation for language and meaning in ML systems. Based in Tallinn, he’s known for pragmatically blending automation, reproducibility and flexibility to accelerate model delivery in regulated enterprise settings.
13 years of coding experience
18 years of employment as a software developer
MSc Data Science, MSc Data Science at KTH Royal Institute of Technology
MSc Data Science, MSc Data Science at Eindhoven University of Technology
BSc Computer Science, BSc Computer Science at University of Southampton
BA Semiotics and Cultural Studies, BA Semiotics and Cultural Studies at University of Tartu