Dr Nils Mattias Hansson
Computer Vision and AI Specialist
Experience
Freelance consultant
Pixel Dot ML Solutions, Copenhagen, Denmark
​​Sep 2023 - present
Skills
Computer Vision
Machine Learning
Mathematical modelling
Statistics
​
Senior Research Scientist
Sartorius Stedim Data Analytics, Malmoe, Sweden.
Aug ​2022 - Feb 2023
Projects in 3d microscopy data and infrastructure. Tasks included development of an ETL pipeline with the following proprieties:
-
input: fluorescent microscope data, output: optimised annotation of cells.
-
extraction of metadata to Cosmos DB, cleaning and cropping of image data.
-
data is fed into third-party annotation platform, where annotation are refined using dockerised ML models.
-
Computer Vision models are retrained using refined annotation data.
-
built using Python on Azure cloud platform using Azure functions, specifically Durable (stateless) functions.
​
Data Scientist
Neurons Inc, Copenhagen, Denmark
​Sep 2021 - June 2022
Leading work in applied neuroscience including creation of neurometrics and AI models for sentiment prediction. Supervision of statistical modelling work of data analysts.
Languages
Python
R
SQL
Linux
Bash
AWS
Azure
GCP
​
​
Senior Mathware Engineer
Sioux Technologies, Eindhoven, The Netherlands
​Oct 2020 - Sep 2021
Senior consultant in Computer Vision and Machine Learning. Projects included development of computer vision applications in biotech serving
an industry leading player.
Data Scientist
Pixel Data Science Solutions, Rotterdam, The Netherlands.
​Oct 2019 - Oct 2020
Development of platform agnostic machine learning solutions for universities and companies.
Company discontinued due to start of employment at Sioux Technologies.
Data Scientist
Dept Digital Marketing, Rotterdam, The Netherlands.
​Oct 2018 - Oct 2019
Development and deployment of machine learning models for analysis of digital marketing data
-
input: raw sales data, output: probabilistic customer journey graph.
-
using Markov model to predict behavior from sales entry point.
-
built using Python on Azure Cloud Platform.
​
Creation and deployment of energy consumption model for houses in the NL market.
Scientific Programmer
Erasmus Medical Center, Rotterdam, The Netherlands
​Oct 2015 - Oct 2018
Construction of computer vision methods for association studies with the following properties:
-
input: MRI scans of human knees, output: volume of knee structures.
-
segmenting knee structures, extracting bio-markers, determining association between selected SNPs and biomarkers.
-
built using Python on in-house server infrastructure, due to privacy requirements of medical data in the Netherlands.
-
Association study analysis coded in R.
Post-doctoral Researcher
University of Copenhagen, Copenhagen, Denmark.
​Feb 2013 - Oct 2015
Responsible for development of algorithm (based on sparse methods and dictionary learning) for detection of defects in x-ray images of wood planks to optimize performance in furniture production in the industrial project TreeDFurniture.
Education
Ph.D. Mathematics, Lund University, Sweden
​2006 - 2013
Thesis: "Statistical Segmentation and Registration of Medical Ultrasound Data.". Development of methods for recovering structures in various forms of ultrasound images.
M.Sc. Mathematics, Linkoping University, Sweden
​2003 - 2005
Thesis: "Numerical Experiments with Femlab to support mathematical research". Mathematical imaging of the solutions to the infinity Laplace equation.