Lars Hertel
Senior Staff Machine Learning Engineer at LinkedIn
San Jose, California, United States
Summary
👤
Senior🎓
Top SchoolLars Hertel is a Senior Staff Machine Learning Engineer in San Jose with a decade of experience building production-grade recommender and ranking systems at LinkedIn. He led the adoption of Transformer- and LLM-based user-history models and founded the Feed’s generative recommender initiative, taking projects from prototype to MME-scale impact while driving cross-org execution and online debugging. Lars combines end-to-end ownership—data pipelines, training, inference, and A/B testing—with research rigor, coauthoring KDD publications (including LiRank) and inventing an amortized inference algorithm now reused across teams. His work has repeatedly produced measurable true-north gains (e.g., +0.21% and +0.25%) and inspired other verticals to ship similar innovations. A PhD-trained statistician, he pairs deep probabilistic thinking with practical systems craftsmanship and a track record of mentoring and elevating junior engineers. Notably, he rebuilt LinkedIn’s Feed ranking stack with a large sequential ranker and drove the company’s first DCNv2 and Transformer recommender deployments.
10 years of coding experience
5 years of employment as a software developer
Master of Science (MS), Statistics, Master of Science (MS), Statistics at Imperial College LondonBachelor of Science (BSc), Physics, Bachelor of Science (BSc), Physics at Queen Mary University of LondonUniversity of California, Irvine
English, German
Github Skills (17)
python9
visualize9
scale9
gpy9
optimization9
gaussian8
hyperparameter-optimization8
machine-learning8
pytorch8
data-science8
caffe7
tensorflow7
deep-learning7
neural-network5
sparse4
Programming languages (5)
Github contributions (5)
LarsHH/LarsHH.github.io
Jul 2016 - Jul 2021
Contributions:4 PRs, 155 pushes, 6 branches in 5 years
sherpa-ai/sherpa
Apr 2017 - Apr 2019
Hyperparameter optimization that enables researchers to experiment, visualize, and scale quickly.
Contributions:6 releases, 92 commits, 93 PRs in 2 years
pythondata-sciencehyperparametervisualizeoptimization
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.