Dmitry Kravchenko

Software Engineer & Architect · AI Systems · Physicist by training

Profile

Physicist by education (MSc, Moscow State University — numerical modelling of nucleus–nucleus collisions), software engineer and architect by career, and AI systems practitioner by recent work. 25+ years of hands-on development across distributed systems, machine learning, and applied AI, with a repeated track record of turning early-stage research into working software.

Experience

Booking.com — Software Engineer, AI & Data Aug 2021 – Present

Backend & AI platform engineering. Netherlands.

MicroArt — Embedded & Mobile Developer Aug 2020 – Aug 2021
  • Linux drivers and daemons for battery-charger hardware (C, CMake, ARM cross-compilation).
  • Migrated Android, iOS and Windows apps to new hardware revisions.
  • Built a Jenkins pipeline for automated Raspberry Pi cross-compilation.
Yandex — Big Data Developer Apr 2019 – Jul 2020
  • Cut machine time 10× processing a 300-billion-record database.
  • Reduced one map-reduce stage from 5h to 1h on a 1-trillion-record dataset.
  • Automated image-feature computation (neural + image-to-text) for a main-page widget.
Presenso — Machine Learning Engineer (remote) Dec 2017 – Apr 2019
  • PoC applying deep models (LSTM, GRU, conv) to time-series data, showing their advantage.
  • Automated model training and hyperparameter search across Matlab and Hadoop/Spark.
EnergentGroup — Python GIS & ML Programmer (remote) May 2017 – Aug 2018
  • Built a Keras VGG-like deep CNN for change classification — 95.8% F-beta.
  • Sentinel-2 training-data collection pipeline; AWS Lambda image processing on S3.
Cireca Ltd — Java Programmer (remote) Mar 2016 – Dec 2016
  • Laboratory information system for data mining on medical-image recognition.
  • Image-analysis module for supervised learning over 10,000+ images.

Technical Skills

AI / ML
PyTorch, TensorFlow, Keras, LLM APIs, LangChain, RAG pipelines, evaluation frameworks, Jupyter
Languages
Java, Python, C/C++, C#, Kotlin, JavaScript/React, Scala, Matlab, SQL
Infrastructure
AWS (S3, DynamoDB, Lambda), Kubernetes, Docker, Kafka, MySQL, MongoDB, GitLab CI/CD, Grafana
Agentic tooling
Claude Code, Cursor, Windsurf, MCP servers
Scientific
Numerical simulation, high-performance pipelines, quantum mechanics & applied maths foundation

Selected Side Projects

Education & Languages

M.Sc., Physics — Moscow State University. Thesis: computer modelling of nucleus–nucleus collisions.

English — fluent · French — reading · learning Mandarin & Dutch.