Sergey Karayev
Sergey Karayev is a computer scientist and software builder whose work spans computer vision, deep-learning frameworks, AI-assisted grading, and coding-agent tools. He did undergraduate research at the University of Washington and completed a computer-science PhD at UC Berkeley in 2014, focused on computer vision. In 2014, he co-authored Caffe, a framework for deep learning; the team received the Everingham Prize in 2017. He helped launch Gradescope as an AI-assisted grading platform, which Turnitin acquired in 2018; by 2020, he led STEM AI work there and co-developed Full Stack Deep Learning courses and a 2023 LLM bootcamp. As of October 2026, he is a co-founder of Superconductor, which builds a shared AI workspace for teams and coding agents.
- Now at
- Superconductor
- Works
- 32
- Citations
- 16,210
- h-index
- 12
Publication numbers from OpenAlex, as of 15 October 2026.
Timeline
-
By 2014
A computer-vision doctorate
He does undergraduate research at UW Seattle before completing a computer-science PhD at UC Berkeley in 2014, focused on computer vision.
-
2014–2017
The Caffe team’s Everingham Prize
He co-authors Caffe, a BSD-licensed C++ framework for training and deploying convolutional neural networks, in 2014. The Caffe team receives the Everingham Prize in 2017 for its open-source deep-learning framework.
-
2014–2018
Co-founding Gradescope
With Arjun Singh and Pieter Abbeel, he develops Gradescope from an AI course’s grading work and forms a company to launch it in 2014. Turnitin acquires the AI-assisted grading platform in 2018.
-
2018–2023
AI for STEM and LLM courses
By 2020, he is Head of STEM AI at Turnitin and writes about design principles for AI-assisted grading. He co-develops Full Stack Deep Learning courses offered from 2018 to 2022 and teaches a lecture in its 2023 LLM Bootcamp.
-
As of October 2026
Building Superconductor
He is a co-founder of Superconductor, where he builds a shared AI workspace for teams and coding agents.
Sources and credits
Photo: We have no published photo of Sergey Karayev yet. Send us a free-licensed photo.
- Gradescope: Taking the pain out of grading
- Sergey Karayev | Home
- Caffe: Convolutional Architecture for Fast Feature Embedding
- Turnitin Acquires Gradescope
- Turnitin Tech Talk: Design Principles of AI-Assisted Grading
- What’s Next? - The Full Stack
- Sergey Karayev | Speaker | ASU GSV Summit
- What is Multiplayer AI? | Superconductor Blog