AI Engineer
Я интересуюсь данными, машинным обучением и тем, как технологии можно применять для решения реальных задач.
Мне нравится разбираться в сложных системах, искать закономерности в данных и превращать идеи в работающие решения. Сейчас я активно развиваюсь в направлении AI и Data Science, стараюсь больше практиковаться, участвовать в проектах и расти через общение с людьми из этой сферы.
Beles AI
Nov 2025 — Feb 2026Built computer vision pipelines and supported production integration of three models, including face recognition and identity matching.
• Built end-to-end CV pipelines for data preparation, training, inference, and evaluation; supported integration of 3 CV models into production workflows. • Worked on face recognition and identity matching under real-world constraints: small faces, camera angle variation, blur, lighting changes, and low image quality.
Armeta AI
Oct 2025 — Feb 2026Developed an LLM-based document processing system and retrieval pipelines covering more than 5,000 documents.
• Developed an LLM- and agent-based document processing system for high-volume workflows with fast response requirements. • Built retrieval/indexing pipelines over 5K+ documents, including parsing, chunking, embeddings, vector search, and context construction. • Implemented backend APIs for document extraction, analysis, structured outputs, and service integration.
Adata.kz
Jun 2024 — Aug 2025Built ML/LLM services and RAG pipelines, and improved database workloads by 25% while ETL workflows processed 500K+ records daily.
• Built production ML/LLM inference services using FastAPI and gRPC, integrating AI workflows with internal APIs, databases, and backend systems. • Developed RAG pipelines with FAISS and Qdrant, supporting 40K+ indexed chunks for semantic retrieval and AI-enabled search. • Designed ETL workflows with Airflow and Talend, processing 500K+ records/day; optimized PostgreSQL/ClickHouse workloads by 25%.
SDU Technopark
Sep 2023 — Jul 2025Assisted more than 10 ML/NLP research and lab projects with data preparation, experimentation, and evaluation.
• Supported 10+ ML/NLP research and lab projects through dataset preparation, experimentation, model evaluation, tooling, and implementation assistance.
Bizdin AI
Jun 2023 — Oct 2023Prepared 40,000 training samples and fine-tuned BERT and LLaMA-based models, improving validation results by 8%.
• Prepared 40K training samples for LLM fine-tuning; fine-tuned BERT and LLaMA-based models with LoRA/PEFT, improving validation results by 8%. • Improved preprocessing and training workflows to reduce unnecessary compute usage and support reproducible experiments.
Astana IT University
2025 — 2027SDU University
2021 — 2025Built a document QA chatbot with LangChain, FAISS, local Mistral, and OpenAI fallback.
Prepared bilingual QA data and fine-tuned LLaMA 2 using PEFT/LoRA.
Added observability and deployment support.
Data Science Hackathon · 2023
Data Science Hackathon · 2024
