Andrej Karpathy
@karpathy
#3 on Gitdex·Top 0.18% (est.)
I like to train Deep Neural Nets on large datasets.
More information is visible only for Pro subscribers.
Andrej Karpathy
AI-Powered Insights
Distinguished Engineer / FellowOpen to opportunities: UncertainAndrej Karpathy is a world-renowned AI researcher and educator specializing in deep learning and computer vision, best known for his pivotal roles at OpenAI, Tesla, and Stanford University where he led the AI team for Autopilot and co-developed foundational technologies like GPT-2 and nanoGPT. With a PhD from the University of Toronto under Geoffrey Hinton, he has authored influential open-source projects such as micrograd and llm.c, amassing hundreds of thousands of GitHub stars and millions of views on his educational YouTube series like Neural Networks: Zero to Hero. Currently an independent researcher after leaving OpenAI in 2024, he focuses on making AI accessible through tutorials, tools, and his personal site karpathy.ai.
$1,500,000 – $3,000,000
Karpathy demonstrates exceptional depth in deep learning frameworks like PyTorch and TensorFlow, with hands-on expertise in training massive models on large datasets as evidenced by his work on GPT-scale architectures and Tesla's vision-based autonomy stack. His breadth spans low-level implementations in C (e.g., llm.c) to high-level educational tools in Python and JavaScript, showcasing polyglot proficiency including Go and C++. Most impressive is his ability to distill complex concepts into minimal, efficient codebases like nanoGPT and micrograd, influencing millions of developers.
nanoGPT
A lightweight, from-scratch implementation of GPT in PyTorch for training character-level language models, created by Karpathy to teach transformer mechanics and achieve state-of-the-art results on tiny datasets.
micrograd
A tiny autograd engine in 150 lines of Python that powers backpropagation for neural nets, authored by Karpathy as an educational tool to explain core ML concepts from the ground up.
llm.c
An efficient inference engine for large language models written entirely in C, developed by Karpathy to demonstrate high-performance LLM serving without Python dependencies.
Reference specific projects like nanoGPT or his Tesla Autopilot work to show genuine familiarity, and align your outreach with his passion for AI education and efficient LLM implementations by mentioning opportunities in foundational AI research or open-source tooling. Use his public Twitter (@karpathy) for an initial low-pressure mention before emailing, as he is highly active there and engages with technical discussions. Emphasize mission-driven roles in autonomous systems or AGI, avoiding generic pitches given his high profile and selectivity.
Pro: full profile
You're already browsing the index for free. Subscribe once to unlock GitHub activity stats, contact details, market signals, and AI insights on every profile — including karpathy.
Already subscribed? Sign in