Summary
Ilya Gavrilov is a Machine Learning Engineer based in London with ~8 years of adtech experience building end-to-end solutions that run on billions of daily events and hundreds of terabytes of data. He specializes in low-latency, resource-constrained production models (1 CPU core, 5–10ms inference) and has driven measurable gains such as a 30% improvement in bid landscape prediction accuracy and a 20% increase in win rates for CTR/ATR systems. His work spans applied research, experiment design, scalable data pipelines, A/B testing and operational maintenance during major infra migrations. Comfortable across Python, TensorFlow/Keras, LGBM and Hadoop/PySpark, he also brings a strong background in feature engineering and unsupervised deep learning for user vectorization and URL/NLP feature extraction. Colleagues would note he prefers to "pronounce data, not data"—a concise culture of treating metrics as first-class citizens when turning research into production.
8 years of coding experience
5 years of employment as a software developer
Bachelor of Computer Science, Major: Machine Learning and Applications, 8.42 / 10, Bachelor of Computer Science, Major: Machine Learning and Applications, 8.42 / 10 at Higher School of Economics