Andrey Lukyanenko

I'm Andrey Lukyanenko. I work as a Machine Learning Engineer at Meta in London.

I didn't start in tech. My background is in economics and ERP consulting, where I learned to think in systems, communicate with different stakeholders, and understand business problems from the ground up.

Since 2017, I've worked across telecom, banking, medical AI, ride-hailing, and big tech on problems like churn prediction, fraud detection, recommendations, video super-resolution, chat-bots, and ads targeting at scale. Along the way, I became a Kaggle Competition Master and Notebook Grandmaster (reaching #1 in the global notebook ranking), a Google Developer Expert, and I've given 20+ public talks on ML topics.

Meta, London Kaggle Competition Master Notebook Grandmaster Google Developer Expert

What I write about

This site is where I share what I learn. I write detailed reviews of ML papers (190+ so far), posts on applied ML and career, and occasionally about language learning and knowledge management systems.

Latest Paper Reviews

All reviews →
Mar 16, 2026
Collaborative Reinforcement Learning: Why HACRL Trains Models in Teams Instead of Isolation
HACRL proposes a new paradigm for reinforcement learning - instead of training models in isolation, multiple agents collaborate by sharin...
rl
Feb 23, 2026
Beyond Positional Bias: How DroPE Unlocks Zero-Shot Long Context in LLMs
A review of DroPE, a simple but counterintuitive method that extends LLM context length by dropping positional embeddings at inference an...
llm
Feb 16, 2026
Kimi k2.5 Review: Native Multimodality and Agent Swarms at 1 Trillion Parameters
A deep-dive review of Kimi K2.5, a next-generation open multimodal model that combines native vision-language training with parallel agen...
llm

Featured Work

Job Search & Interview Prep as MLE
Honest account of my 2024 job search process, preparation strategy, and lessons learned on the way to Meta.
career interview
Kaggle: EDA and Models (1000+ upvotes)
Feature engineering and fraud detection modeling for IEEE-CIS. One of the most upvoted Kaggle notebooks.
kaggle fraud detection lightgbm
Fraud Detection via Face Recognition
How we built a face-recognition system to catch duplicate driver accounts at Careem — 5% reduction in fraud.
production ML anti-fraud computer vision
Crafting My Personal Website
Jekyll, GitHub Pages, and a custom domain — still one of my most-read posts.
tutorial web development

See all projects →

Beyond Work

12 Years with Anki
How spaced repetition shaped my learning of German, Spanish, and Japanese over a decade.
languages anki
Two Years of Language Practice
An update on my polyglot journey — what worked, what didn't.
languages
KindleVocabToAnki
A tool I built to extract Kindle vocabulary and import it into Anki for language learning.
languages tool
Fun facts
  • I'm an economist by education
  • I have read "War and Peace" three times
  • At some points in my life, I have studied taekwondo, fencing, and chess
  • I have Lean Six Sigma Green Belt certificate
  • I have more than 600k total XP on Duolingo
  • I passed B2 for German on Deutsche Welle in 2013
  • I have an old blog about learning languages
  • I passed JLPT3 in J-CAT test in August 2014
  • I'm one of the authors of a telegram channel Data Science by ODS.ai
  • I love origami
  • Favorite fantasy authors: John Bierce, Brandon Sanderson, Will Wight