Kirill Gelvan

LLM Post-Training Engineer at JetBrains

Munich, Bavaria, Germany
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

🤩
Rockstar
🎓
Top School
Kirill Gelvan is an AI agents researcher and NLP engineer with eight years of experience and a focused three-plus years in production NLP and RAG systems. Based in Munich, he built scalable retrieval-augmented pipelines and error-classification automation that reduced unethical responses by 35% and extended chatbot memory threefold while increasing ARPU. At JetBrains he developed an IDE Debugging-Agent and explored agent distillation and implicit context condensation to tackle software-engineering benchmarks. He combines hands-on model distillation, quantization and vector DB scaling with teaching experience—creating and delivering multiple university and Coursera courses—bringing both practical product impact and clear technical communication to cross-functional teams. Notably, he emphasizes team cohesion and confident leadership alongside technical depth, making him effective at turning research ideas into reliable engineering features.
code9 years of coding experience
job2 years of employment as a software developer
bookMaster of Science Mathematics in Data Science; Post-ChatGPT era, Master of Science Mathematics in Data Science; Post-ChatGPT era at Technical University of Munich
bookBachelor's degree 𝐅𝐚𝐜𝐮𝐥𝐭𝐲 𝐨𝐟 𝐂𝐒, Bachelor's degree 𝐅𝐚𝐜𝐮𝐥𝐭𝐲 𝐨𝐟 𝐂𝐒 at Higher School of Economics
stackoverflow-logo

Stackoverflow

Stats
1reputation
0reached
0answers
0questions
github-logo-circle

Github Skills (63)

transformers10
pytorch10
python10
machine-learning10
multimodal10
llm10
audio10
deep-learning10
natural-language-processing10
transformer10
speech-recognition10
nlp10
reinforcement-learning9
machine-learning-models9
depth-estimation9

Programming languages (7)

C++ShellTeXJavaScriptJupyter NotebookPythonKotlin

Github contributions (5)

github-logo-circle
Kirili4ik/HRL-taxi

Jul 2018 - Nov 2019

Solution for Taxi env using HRL (Hierarchical reinforcement learning) (2018)
Contributions:25 commits, 2 PRs, 24 pushes in 1 year 4 months
reinforcement-learninghierarchical-reinforcement-learningmaxqrlpython3
Kirili4ik/code2vec

Apr 2020 - Aug 2021

code2vec for Python 3 made for NL2ML project
Contributions:22 commits, 10 pushes, 1 branch in 1 year 4 months
pythoncode2vecembeddingspython3
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.
Request Free Trial