Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Mikhail Isakov, an ML engineer with 3+ years of experience across academia and industry, specializing in computer vision and NLP. I've delivered end-to-end AI solutions—from data collection and preprocessing to model training and deployment—in OCR, RAG, and real-time tracking systems. I also work in data engineering for pipelines, databases, and real-time processing. With a software development background and a computer science education, I stay on top of AI advances, follow research trends, and enjoy hackathons.…I'm Mikhail Isakov, an ML engineer with 3+ years of experience across academia and industry, specializing in computer vision and NLP. I've delivered end-to-end AI solutions—from data collection and preprocessing to model training and deployment—in OCR, RAG, and real-time tracking systems. I also work in data engineering for pipelines, databases, and real-time processing. With a software development background and a computer science education, I stay on top of AI advances, follow research trends, and enjoy hackathons.

Mikhail Isakov

Data Scientist, AI Engineer, AI Developer, +2





I'm Mikhail Isakov, an ML engineer with 3+ years of experience across academia and industry, specializing in computer vision and NLP. I've delivered end-to-end AI solutions—from data collection and preprocessing to model training and deployment—in OCR, RAG, and real-time tracking systems. I also work in data engineering for pipelines, databases, and real-time processing. With a software development background and a computer science education, I stay on top of AI advances, follow research trends, and enjoy hackathons.…I'm Mikhail Isakov, an ML engineer with 3+ years of experience across academia and industry, specializing in computer vision and NLP. I've delivered end-to-end AI solutions—from data collection and preprocessing to model training and deployment—in OCR, RAG, and real-time tracking systems. I also work in data engineering for pipelines, databases, and real-time processing. With a software development background and a computer science education, I stay on top of AI advances, follow research trends, and enjoy hackathons.

Available to hire

I’m Mikhail Isakov, an ML engineer with 3+ years of experience across academia and industry, specializing in computer vision and NLP. I’ve delivered end-to-end AI solutions—from data collection and preprocessing to model training and deployment—in OCR, RAG, and real-time tracking systems.

I also work in data engineering for pipelines, databases, and real-time processing. With a software development background and a computer science education, I stay on top of AI advances, follow research trends, and enjoy hackathons.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Work Experience

ML Engineer at SberBank

October 1, 2025 - October 15, 2025

Trained and pruned multimodal LLM for formula recognition in scientific documents, achieving a BLEU score of 0.93. Fine-tuned end-to-end OCR and layout models for bank-specific documents, reaching SOTA in the domain. Designed a RAG system, fine-tuned embedder and reranker models via SFT raising mAP to 0.84, which improved customer support and answer quality. Deployed ML models with Docker, Kubernetes, and CI/CD pipelines, ensuring reliable, scalable, and efficient production.

ML Engineer at Bachelor’s Thesis - Real-Time Tracking with PTZ cameras

June 1, 2024 - October 15, 2025

Designed and deployed a real-time MCMT tracking system for stage capture, improving tracking accuracy by 30% and enabling robust handling of crossing-path scenarios. Containerized ML models and PTZ control pipeline with Docker, integrating FFmpeg/GStreamer streams and VISCA/ONVIF protocols, which improved deployment scalability and reduced setup time by 50%.

Teaching Assistant at Higher School of Economics University

January 1, 2024 - October 15, 2025

Taught 20 students and graded 140 practical homework assignments in a machine learning course.

ML/Data Engineer at Safe Transport Innovation Center

October 1, 2023 - October 15, 2025

Built scalable Flask services for automated social media data scraping (200+ GB), leveraging RabbitMQ for distributed task management, MongoDB for storage, and Grafana for real-time monitoring. Analyzed a large-scale media resource graph (25.9M posts, 278K channels, 929K links) with clustering, centrality, and temporal analysis, revealing key community structures and their evolution. Trained MiniLM for topic classification of social media posts, achieving a Weighted F1-score of 0.91. Integrated NER models into data pipelines, streamlining downstream data analysis.

Software Developer at DaisyKnit

October 1, 2021 - October 15, 2025

Built a customer feedback chatbot in Telegram with FastAPI and PostgreSQL, integrating CI/CD pipelines.