Sergey Kolesnikov
Head of AI Research
Tl;dr
I am Head of AI Research at Tinkoff, an AI-first fintech company. Our team is focused on the CV, NLP, RL, and RecSys domains to advance the AI field and create novel technical solutions. We also frequently launch students' projects through TLab.
My own line of research is mainly focused on sequential decision-making under uncertainty and ML robustness. Besides that, I help with the Tinkoff.AI meetups organization and advise the MLE profession within Tinkoff.
In my spare time, I actively contribute to open-source projects and lead the development of Catalyst, an open-source PyTorch framework for deep learning research and development. I also recently launched Animus as a framework-less alternative to organize research experiments.
Short notes in Decision Making in the Wild.
Posts
Talks
Publications
Projects
CORL
High-quality single-file implementations of SOTA Offline RL algorithms: AWAC, BC, CQL, DT, EDAC, IQL, SAC-N, TD3+BC
Animus
Minimalistic framework to run machine learning experiments
4th place solution for the SIGIR 2021 challenge.
4th place solution for the SIGIR 2021 challenge.
ETNA
Time Series Library
Kittylyst
A tiny Catalyst-like experiment runner framework on top of micrograd.
Deep Learning with Catalyst
Advanced deep learning course
ML workflow
Checklist and questions to make ML right.
Catalyst codestyle
Accelerated Python code formatter
RL Intro
Brief introduction to Reinforcement Learning and Recommender Systems.
Alchemy
Experiments logging & visualization
NeurIPS RL 2019 challenge
2nd place solution for NeurIPS RL 2019 challenge
Reaction
Convenient deep learning models serving
Catalyst detection
Object detection with Catalyst
Catalyst segmentation
Image segmentation with Catalyst
Catalyst classification
Image classification with Catalyst
Catalyst.RL
Distributed framework for RL research
Catalyst
Accelerated deep learning research and development
NeurIPS RL 2018 challenge
3rd place solution for NeurIPS RL 2018 challenge
OpenAI Retro Contest
4th place solution for OpenAI Retro Contest.
NIPS RL 2017 challenge
3rd place solution for NIPS RL 2017 challenge