Avatar

Magnus Berg Sletfjerding

Machine Learning Engineer

Corti

Nørrebro Klatreklub

Biography

I am currently working as a Machine Learning Engineer at Corti.ai

Otherwise I spend my time managing the web content of Nørrebro Klatreklub in Copenhagen, Denmark.

I also like herbal teas and endurance training.

Interests

  • Probabilistic Generative Models
  • Image Analysis and Segmentation
  • Responsible Governance & International Relations
  • Rock Climbing and Sailing

Education

  • M.Sc. in Nanoscience / Biophysics, 2022

    University of Copenhagen

  • B.Sc. in Nanoscience, 2018

    University of Copenhagen

Skills

Python

I use Python on a daily basis to analyze microscopy images, time series, and build predictive models and software for our research group

Teaching

I have volunteered as a trainer and facilitator, and have taught courses in Nanobiology and Spectroscopy for B.Sc. students at UCPH.

Statistics

My research is based on the use of statistical techniques and models to understand protein structure and function

Work Experience

 
 
 
 
 

Data Engineer [Full-Time]

Corti

Mar 2022 – Present Copenhagen, Denmark

Corti builds intelligent machines for augmented medicine.
I contributed in the following ways:

  • Independently scoped and built a gRPC microservice adapter for real-time data streaming and conversion to Corti’s online detection service, decreasing latency between services by 70%.
  • Main architect of a Data Warehouse solution to allow customers and internal stakeholders access to their own data, unlocking insights for 30% call reduction time.
 
 
 
 
 

Junior Machine Learning Engineer [Part-Time]

Corti

Mar 2021 – Feb 2022 Copenhagen, Denmark

Corti builds intelligent machines for augmented medicine. Our AI supports medical personnel on making decisions in real time during medical emergencies, combining bleeding-edge machine learning research and real-world data from emergency calls. I worked on the following types of projects:

  • Scoped and built a full-stack end2end testing framework for Microsoft Azure cloud deployments.
  • Create and maintain end2end CI/CD Microsoft Azure pipeline for scalable Dockerized using Google developed gRPC-based microservices with Tensorflow models built-in.
 
 
 
 
 

Intern Machine Learning Engineer [Full-Time]

Corti

Aug 2020 – Mar 2021 Copenhagen, Denmark

Corti builds intelligent machines for augmented medicine. Our AI supports medical personnel on making decisions in real time during medical emergencies, combining bleeding-edge machine learning research and real-world data from emergency calls. I worked on the following types of projects:

  • Building and testing REST-APIs for SQL and NoSQL (Mongo, ElasticSearch) databases in Azure Functions.
  • Constructing and deploying Data Pipelines with Azure Data Factory
  • Automating cloud deployments of Dockerized machine-learning based microservices.
  • Develop and deploy machine-learning models for real-time NLP tasks using TensorFlow.
  • Improved data upload pipeline to pSQL databases by 98% from ~3 days to 15 minutes.
  • Migrate Jenkins-based CI/CD hosted in Amazon AWS to Microsoft Azure Pipelines
 
 
 
 
 

Student Assistant [Part-Time]

MAN Energy Solutions

Sep 2019 – Aug 2020 Copenhagen, Denmark

Projects:

  • U-NET segmentation of microscopy images to quantify wear of piston ring coatings.
  • Characterizing graphite defects in cast iron by morphology analysis.
 
 
 
 
 

3+5 Ph.D. Student [Full-Time]

The Hatzakis Lab, University of Copenhagen

Feb 2019 – Apr 2021 Copenhagen, Denmark

Use of a variety of statistical models (ConvNets, LSTMs, Hidden Markov Models) on diverse biological datasets to streamline current data processing pipelines and determine hidden trends in single-molecule enzymology.
Projects:

  • Automating data analysis of smFRET traces in DeepFRET.
  • Simulation of CYPOR conformational dynamics.
  • Statistical analyses of multiplexed liposome assays in SPARCLD
 
 
 
 
 

Seasonal Staff / Dreng [Full-Time]

Vangshaugen Tourist Cabin

Jul 2018 – Aug 2018 Grødalen, Norway

Maintaining and running a Tourist Cabin for the Norwegian Tourist Association. Responsibilities:

  • Tenant relations
  • Serving at all meals (>40 people)
  • General maintenance of the property
 
 
 
 
 

Summer Intern in Procurement, 2 seasons [Full-Time]

BioMar A/S

Jul 2016 – Aug 2017 Trondheim, Norway

Responsibilities:

  • Updating SharePoint, Excel, and M3 databases with biochemical analysis results for shipments
  • Maintaining Excel spreadsheets and assisting in preparing Q3 reports
  • Taking minutes at multilingual phone meetings in the Back Office department

Volunteer Causes

I’ve been involved in a series of volunteer causes. Here are some of them.

Student Contact and Head of Copenhagen

I got involved in Tekna in the spring of 2018. As a Student Contact, I had the opportunity to organize events and gatherings for Norwegian students within scientific and technical subjects. Later, as the Head of the Copenhagen Branch, I have had the pleasure to oversee a series of different events for our volunteers and members.

Founder

When I moved to Denmark, my friend Endre and I founded The European Youth Parliament Denmark, to provide a platform for Danish youth to express themselves, and debate global affairs. I had the great pleasure to serve as President of the Board, and be part of the quintessential starting phase for the organization.

Editor, Chairperson, Media Team Member, Organizer

During my time at Trondheim Cathedral School, I participated in a debate organized by the European Youth Parliament Norway. Since, I participated in many different events, both as a participant and as a volunteer, and met a whole bunch of great people across Europe.

Recent Publications

Quickly discover relevant content by filtering publications.

Generative Modelling of Sequential Data

Autoregressive convolutional neural networks such as WaveNet are powerful models that have recently achieved state-of-the-art results …

Across kingdom biased CYP-mediated metabolism via small-molecule ligands docking on P450 oxidoreductase

Combining single molecule FRET and single turnover studies with cell studies and docking simulations we advance our understanding on the intricate role of conformational dynamics on activity and specificity, and eventually how pathogenic mutations and small molecule ligand interactions control metabolic disorder and biosynthetic pathways.

DeepFRET

DeepFRET, a new deep learning-based platform for standardized, automated, and unbiased single-molecule FRET data analysis, seeks to lower the threshold for smFRET expertise, allowing for a greater number of scientists to take advantage of this powerful technique.

Talks

The Role of Conformational Dynamics in Protein Function

Enzymes regulate a plethora of vital cellular processes, and aberration in their function can result in countless disease states. …

Accomplish­ments

I like learning new things.

Divide and Conquer, Sorting and Searching, and Randomized Algorithms

Asymptotic (‘Big-oh’) notation, sorting and searching, divide and conquer (master method, integer and matrix multiplication, closest pair), and randomized algorithms (QuickSort, contraction algorithm for min cuts).
See certificate

JavaScript Algorithms and Data Structures

Developed a working understanding of JavaScript and how to program simple algorithms for use in frontends.
See certificate

Instructor 2 – Bouldering, Toprope & Lead Climbing

The Instructor 2 course qualifies for teaching and certifying new climbers in Toprope belaying, Sport Climbing, and belaying Sport Climbing in Denmark. The course is a 15 hour hands-on course with 30 hrs of preparation time, also covering basic safety in structuring bouldering and general training for beginner climbers.
See certificate

Introduction to University Pedagogy

Practiced development and execution of learning goal-based course design and teaching, in a week-long workshop for employees at the Faculty of Science. Covered student activation techniques, and problem-based learning engagement.
See certificate

Ernst & Young Talent Training for Management Consulting

Covered introduction to EY’s style of problem solving and presentation techniques. Six session with topics and instructors from across different service lines of EY’s Advisory branch. My main focus was new digital technologies and their business applications.

Projects

Traffic Simulation

I made this traffic simulation as part of Namiko Mitarai’s Course in Diffusive and Stochastic Processes in 2020

Neural Network in NumPy

To better understand the mechanics of a neural network, I made my own implementation to test code and small examples of neural networks.

Twitter Reply Scraper

I created this scraping tool with BeautifulSoup and Selenium to scrape replies to Tweets for subsequent sentiment analysis.

Probability Slider

I made this tool to visualize different probability distributions. It’s not very extensive but it’s proved very useful when discussing probability distribution while teaching basic statistics.

Get in Touch