Expertise in AI, Machine Learning & Software Development
Jesper Derehag is a freelancer, providing advanced engineering expertise in Machine Learning, AI, and large-scale software systems.
Based in Kungälv, Sweden • Serving Gothenburg, Stenungsund and surrounding areas (or remote)
Get in TouchMy Services
Machine Learning & AI
From ideation and proof-of-concept development to putting AI solutions into production.
Software & Systems Architecture
With 20 years of experience in software development and architecture, I provide technical leadership, architecting new features, and continuously improving complex legacy codebases.
Performance Engineering
Expertise in optimizing system performance, characteristics and low-level systems and libraries, ensuring your applications run efficiently.
ML System Deployment & Training
Helping teams get their machine learning systems from a proof-of-concept to a robust, production-ready state. This includes everything from architecture and pipeline design to deployment and continuous monitoring.
Open-Source Contributions
I am an active contributor to many different open-source projects, including the Linux kernel. I can help you navigate and contribute effectively to the open-source community.
From embedded to cloud
With extensive experience from both embedded, all the way to on-prem custom HW as well as public cloud solutions
About Jesper Derehag
Jesper is a freelancing Tech Lead and Senior Machine Learning Engineer with a career spanning over 20 years in software development, architecture, and research. His journey in machine learning began more than two decades ago with the development of a small embedded computer-vision system and has since spanned everyting from natural-language processing to reinforcement learning and everything in between.
Broad and deep understanding of both ML and system architecture, making him the go-to person for complex technical challenges. With experience from both large enterprises and lean startups, giving him a unique perspective on how to tackle challenges from different angles. This allows him to develop robust and scalable solutions, even with limited resources.
A highly technical ML Engineer, avid open-source proponent and a contributor to many projects, including the Linux kernel.
Highlights
20
Years in AI/ML, SW Development & Architecture
9
Published Papers
6
Patents
Technical Skills
Languages
- Python
- C/C++
- golang
- Assembler (x86, ARM, mips)
- Java
- JavaScript
- C# .NET
- GNU make, cmake
- Bash/csh
Libraries & Frameworks
- PyTorch, TensorFlow
- sklearn, scipy, pandas, numpy
- Spark, Hadoop, Ray, rllib
- Kafka
Platforms & Tools
-
Operating Systems
Linux, *BSD, preempt-rt, RTXC, Contiki, uCLinux
-
Version Control
git, ClearCase, pvcs, cvs, svn, gerrit, gitlab
Talks & Presentations
-
Github Co-Pilot and legal aspects of auto-generated code
Software center | 2023
GitHub Co-pilot and open source systems: legal aspects of auto-generated code
-
ML in Practice - interaction between ML modeling and downstream design decisions
Gothenburg AI Alliance | 2022
ML from an engineering perspective, using a mobility prediction project at Ericsson, focusing on how model selection impacts the final solution with a custom wait-free sparse graph implementation with approximate updates for production use cases.
-
Machine learning in practice - the engineering perspective
RISE - Learning Machines seminar | 2021
A talk focusing on the practical engineering challenges and solutions for real-world ML applications.
-
Lindholmen Software development day | 2020
Examining how AI can be used to improve software development processes and tools.
-
Exploration and evaluation of reinforcement learning in production
Gothenburg AI Alliance | 2019
The typical assertion is that RL does not work in production systems due to the exploratory nature of agents, but is it possible to mitigate some of these assumptions? This talk will be about issues with exploration and evaluation in RL production systems, but also about mitigation in terms of sample-efficiency (for ex. through transfer or distributed/federated learning), safe exploration, and off-policy evaluation.
-
On the challenges in transitioning from embedded to microservices
ICSA/AMS - Keynote | 2017
A keynote discussing the architectural and engineering challenges of migrating from embedded systems to a microservices architecture.
-
Deep reinforcement learning - How DeepMind achieved (super)human ability on playing Atari games
MLDS-GBG meetup | 2017
A technical talk on the algorithms and techniques behind DeepMind's breakthrough in reinforcement learning.
Contact
Ready to start a project or need expert advice? Get in touch with me.
Based in Kungälv, Sweden • Serving Kungälv, Gothenburg, Stenungsund and surrounding areas