Albin Heimerson
Algorithm Developer at Bosch Nordic
Lund, Sweden
Summary
👤
Senior🎓
Top SchoolAlbin Heimerson is an algorithm developer with 11 years of experience, combining a PhD in machine learning and reinforcement learning with hands-on engineering in signal processing and sensor fusion at Bosch Nordic. He specializes in using ML/RL for control problems—most notably smart datacenter control from his doctoral research—and has contributed substantive improvements to the Julia ReinforcementLearning.jl ecosystem, including robust PPO implementations and multi-action-space fixes. With master's degrees in Engineering Physics (Lund) and Computer Science (UC Irvine), he blends strong theoretical grounding with practical tool-building, from in-house testing utilities to production-ready algorithmic components. Colleagues describe him as prolific and curious—his GitHub motto, "Too much to code, too little time," reflects a pattern of exploring diverse solutions and squeezing performance out of complex systems.
12 years of coding experience
Master's degree, Engineering Physics, GPA: 4.94/5, Master's degree, Engineering Physics, GPA: 4.94/5 at Lunds tekniska högskolaUniversity of California, Irvine
Stackoverflow
131reputation
440reached
5answers
0questions
Github Skills (15)
deep-reinforcement-learning10
machine-learning10
ppp10
reinforcement-learning10
julia10
sac9
fluxor8
flux8
debugging6
keras6
arraylist6
socket6
lan6
breakpoints6
java6
Programming languages (12)
Github contributions (5)
JuliaReinforcementLearning/ReinforcementLearning.jl
Mar 2021 - Oct 2021
A reinforcement learning package for Julia
Role in this project:
ML Engineer
ML EngineerContributions:10 reviews, 29 commits, 8 PRs in 7 months
Contributions summary:Albin primarily contributed to the reinforcement learning package by implementing and refining the PPO (Proximal Policy Optimization) algorithm. This included handling multidimensional actions, fixing bugs related to multi-action spaces, and adapting the PPO example for the Pendulum environment. Furthermore, the user replaced the SAC (Soft Actor-Critic) policy network with a Gaussian network, and addressed inconsistencies in wrappers to enhance the package's functionality.
juliareinforcement-learningmachine-learningdeep-reinforcement-learningdeep-q-network
albheim/PoleExploration.jl
Mar 2020 - Nov 2022
Small graphical tool for exploring the behaviour of transfer functions based on their poles and zeros.
Contributions:52 commits, 5 PRs, 24 pushes in 2 years 9 months
transfer-functions
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.