CV

Researcher in applied AI with a PhD in predictive and prescriptive process monitoring and more than ten years of machine learning and AI in industry and research, eight of them in Nordic telecom. Current research covers large language models, agentic systems and their evaluation, with published benchmarks and datasets for Nordic languages.

The most up-to-date version of my CV is always available on my LinkedIn profile.

Research interests

Predictive and prescriptive process monitoring; sequential decision-making in operational processes, including reinforcement learning and model predictive control; evaluation, uncertainty and reliability of AI systems; large language models for Nordic languages; agentic systems for business process management and the human side of working with them.

Experience

Senior Research Scientist

Telenor Research and Innovation, Telenor Group - Oslo, Norway - Oct 2023-present

Applied AI research on large language models, customer service AI, business process intelligence, agentic systems and intelligent automation. Leads own research tracks, contributes to externally funded projects (Horizon Europe INTEND), publishes scientifically and gives strategic briefings to group management. Created Telenor AI Bench, an internal benchmark suite for evaluating language models, and contributed to Telenor’s GenAI strategy work. Methodology lead for process analysis and AI-based automation across group functions.

Senior Data Scientist

Telia Denmark, Business Intelligence and Analytics - Copenhagen - May 2022-Sep 2023

Machine learning and AI for decision support at customer, operational and business level: churn prediction and customer portfolio analysis, workforce planning models for customer service, and field experiments in the customer base. Contributions to Telia Group AI initiatives and the Centre of Excellence for machine learning. Designed and taught a two-day internal machine learning course for the BI team.

Senior Data Scientist

Telia Division X - Stockholm - Jan 2022-May 2022

Automated quality assurance of Crowd Insights, aggregated mobility data from the mobile network, including anomaly detection in time series.

PhD Research Fellow

NMBU School of Economics and Business - Ås, Norway - Aug 2018-Jan 2022

Full-time doctoral position. Research in predictive and prescriptive process monitoring, machine learning and decision support. Teaching and supervision duties (see below). Member of the NORA PhD Advisory Committee.

Data Scientist

Telia Norway, Telia Enterprise and Customer Insights - Oslo - Feb 2018-Dec 2021 (part-time alongside the PhD from Aug 2018)

Development and maintenance of machine learning models for marketing automation and CRM in the enterprise market; churn models, customer analytics and customer value management. Project lead for automated retention activities based on machine learning (2018) and for a common architecture and way of working for machine learning operations (2019).

Research Assistant, Junior Consultant and Consultant

Epinion A/S - Aarhus, Denmark - Nov 2015-Jan 2018

Development of EPI-Q, a system for automated fraud detection in survey data. Quality assurance processes for aviation deliveries (Avinor, Heathrow, Copenhagen Airports, Changi, WizzAir) and analysis projects for Visit Copenhagen, DR, the Danish Ministry of Education, Skånetrafiken, Norsk Tipping and NORAD.

Research Assistant

Department of Management, Aarhus BSS, Aarhus University - Jul 2015-Jun 2017

Qualitative and quantitative analysis for the department’s PhD fellows and management in projects on innovation and management.

Education

Doctor of Philosophy in Economics and Business

NMBU School of Economics and Business, Ås, Norway - 2018-2023

Thesis: Essays on predictive and prescriptive process monitoring (NMBU 2023:21), defended 9 June 2023. Supervisor: Prof. Joachim Scholderer. See research for the essays.

Master of Science in Business Intelligence

Aarhus University, Aarhus BSS, Denmark - 2015-2017

Thesis: A Monte Carlo study of the robustness of artificial neural networks. Supervisor: Prof. Joachim Scholderer.

Bachelor of Science in Economics and Business Administration

Aarhus University, Aarhus BSS, Denmark - 2013-2015

Thesis: User behaviour at Lydmaskinen.dk. Supervisor: Assoc. Prof. Carsten Bergenholtz.

Academy Profession Degree in Marketing Management

Business Academy Southwest (EASV), Sønderborg, Denmark - 2011-2013

Teaching and supervision

Teaching and supervision duties of 569 workload hours at NMBU from 2018 to 2021.

• INN355 Machine learning for business process optimisation (MSc, School of Economics and Business), spring 2019-2021. Process mining, predictive analytics on event logs, time series analysis and decision support. Developed and taught new modules on machine learning on process data.
• INN350 Digitalisation and digital business models (MSc, School of Economics and Business), autumn 2018-2019. Responsible for the analytical and data-driven parts and for supervising project teams in case work with real companies.
• DAT200 Applied machine learning I and DAT300 Applied machine learning II (Faculty of Science and Technology), 2019. Teaching assistant responsible for exercise sessions and project work in practical machine learning and deep learning in Python.
• AOS120 Marketing (BSc, School of Economics and Business), 2019. Teaching assistant.
• Master’s thesis supervision, 2019-2021. Four theses within process optimisation, chatbots, data-driven customer insight and customer behaviour; main supervisor for three.
• Machine learning for BI analysts, Telia Denmark, June 2023. Two-day internal course for the BI team.

Selected projects

• Intelligent Business Process Automation, Telenor (2025-). Methodology lead for analysis and mapping of business processes across group functions, with a playbook and a position paper on robotic process automation versus agentic AI.
• AI Bench and GenAI strategy, Telenor (2024-). Initiator of Telenor’s internal benchmark suite for evaluating language models, productized into a model leaderboard on Telenor AI Factory infrastructure. The open Telenor Nordics corpus came out of the same work.
• BRAGE, Telenor (2024-2025). Project lead and first author of the benchmark for language models on Norwegian customer service dialogues, published at NoDaLiDa/Baltic-HLT 2025.
• INTEND (Horizon Europe, 2024-). Intent-based operation of data infrastructure; business case and research blueprints for the Telenor 5G4Data use case.
• Automated retention and MLOps architecture, Telia Norway (2018-2019). Project lead.
• Automated fraud detection in survey data, Epinion (2016), and discourse analyses for DR and NORAD (2017).

Software and data

• SynBPS: open-source Python framework for simulating business process event logs, distributed on PyPI. Sole developer.
• BRAGE benchmark code.
• Telenor Nordics Customer Service Self-Help Corpus: 1,122 manually validated self-help documents in Danish, Finnish, Norwegian and Swedish (CC BY-NC-SA 4.0).
• Temporal_loss: temporal loss functions for remaining time prediction.

Professional service

• Member, IEEE Task Force on Process Mining (2021-)
• PhD Advisory Committee, NORA - The Norwegian Artificial Intelligence Research Consortium (2021-2023)
• Open-source committee member, Danish Data Science Community (2025)
• Reviewer, Autonomous Agents for Social Good workshop at AAMAS 2025, and Nordic Machine Intelligence (2025)

Talks and presentations

• From prompt to business value: A case study in Telenor Norway. Danish Data Science Community, March 2025.
• The BRAGE benchmark (with Tollef Emil Jørgensen). NoDaLiDa/Baltic-HLT 2025, Tallinn, March 2025.
• Zero-shot classification with large language models. Expert talk, Telenor Learning Center, April 2025.
• The relation between heuristics for business process redesign and predictive/prescriptive process monitoring. PhD trial lecture, NMBU, June 2023.
• Statistical process control and multivariate time series analysis with deep-learning-based anomaly detection. Internal seminar, Telia Division X, 2022.
• Deep learning in practice at Telia Norway. Guest lecture, DAT300, NMBU, 2019.

Entrepreneurship and voluntary work

• Founder and managing director, Studie-24 (2009-2013): production of radio commercials, run alongside my studies.
• Co-founder, U&I Marketing (2012-2013): customer surveys and marketing advice for small businesses.
• Business manager for the band Go Go Berlin, Air Management (2014-2016).
• Voluntary accountant, PHONO Festival, Odense (2015-2016).

Certifications

• Deep Learning Specialization, Coursera / deeplearning.ai (2017-2018): neural networks and deep learning; hyperparameter tuning, regularization and optimization; structuring machine learning projects; convolutional neural networks; sequence models.

Languages

Danish (mother tongue), Norwegian (fluent, daily working language), English (fluent, publishing language), German (basic).

Skills

• Code: Python, R, SQL, SAS, LaTeX
• LLMs and agents: MCP, LLM evaluation, in-context learning, vLLM
• ML and statistics: PyTorch, TensorFlow, scikit-learn, pandas, statsmodels, lifelines
• Process mining: pm4py, Disco, ProM, Celonis
• Platforms and MLOps: Snowflake, BigQuery, PySpark, AWS SageMaker, Vertex AI, Dataiku, Docker, Kubernetes/OpenShift, Git

Publications

See the Research page for a full list of papers with abstracts and links.

Profiles

LinkedIn - Google Scholar - ORCID - GitHub - Hugging Face