Tobias Morville

PhD in Computational Neuroscience
MSc Economics

tobiasmorville@hey.com - tmorville.github.io

Summary

Excellent technical background with more than eight years experience with design, development and deployment of machine learning and statistical models in academia and industry. Currently scaling and running a cross functional team of machine learning experts and developers working aross a variety of products and platforms. In my free time I like to make data art projects and free climb.

Education

2014-2017

PhD in Computational Neuroscience, University of Copenhagen

2010-2010

MSc, Economics; University of Copenhagen

2006-2010

BSc, Economics; University of Copenhagen

Experience

Sept 19 -
Head of Machine Learning, Monstarlab EMEA

Currently leading a small team of machine learning experts, devops, developers and consultants. Responsible for planning and execution, stakeholder management and delivery on all data-centric projects.

• Delivered 20+ machine learning projects across several digital products.
• Day-to-day management and mentoring of a small team of machine learning and data experts.
• Part of the EMEA strategy team for focus on data-centric offerings.

Nov 17 - Sept 19
Senior Data Scientist, 2021.AI

Responsible for technical delivery on several machine learning projects of varying scale. Developed the designed production and Data Science requirements for the AI-platform Grace.

• Developed and deployed more than ten machine learning projects of varying complexity across different industries.
• Cut complexity and improved time-to-delivery by designing and implementing a standard process for the entire data science team.
• Mentoring, teaching and project management.

Nov 17 - May 18
Research Consultant, Danish Research Centre for Magnetic Resonance

Part-time research consultant on experiments that started during my PhD.

• Designed and executed a major experiment with more than 40 participants and 100+ fMRI scans.
• Management of research assiatants, subjects, hardware and data.
• Built pipeline for cleaning and analysis of large amounts of fMRI data.

Nov 14 - Oct 17
Ph.D. Fellow, University of Copenhagen

Developed a biologically plausible model for neural computation. Using predictions from this model, I designed and executed several experiments aimed at falsification of those hypothesis.

• Developed a novel approach to reinforcement learning inspired by statistical physics.
  
• Designed and executed three major experiments resulting in over 100 brain scans and 150 behavioural
• First author of three (soon four, 2021) peer-reviewed scientific articles.
• Co-authored more than ten publications.

Technical expertise

Tech stack

Everything Python (numpy, scipy, scikit, tensorflow, pytorch, xgboost, etc.). Docker, AWS Lambda & ECS or GCP for deployment. Experience with SQL, Mongo, HDFS as well. Novice in Rust.

Methods

Linear, non-linear and deep learning methods for regression and classification across tabular, text and image data. Expert knowledge of reinforcement learning and contextual bandits. Some experience with computer vision and numerical optimization as well.

Extracurricular subjects

Community & outreach

• Offical contributor to TensorFlow, Serverless & Society for Deep Reinforcement Learning.
• Top 20% contributor to Stack Overflow (2020).
• Founder of the “Copenhagen Head/Lead Machine Learning Group”.
• Notes and posts on Monstarlab Engineering blog and my own Github site.
• Member of the Culture Club responsible for events and social gatherings at CPH office.

Hobby projects

• Data art projects using GAN, SR, frame-to-frame prediction and various signal processing techniques for audio synchronization.
• Built a system that achieves ~3% MAPE on estimates for all apartment prices in Copenhagen.

Languages

• Danish (native speaker)
• English (native speaker)