Data engineer (junior)

Описание

Adil Tursynali Almaty, Kazakhstan | +7-778-417-89-38 | tursinaliadil@gmail.com | Telegram: @Joke_m543 Data Analyst | Machine Learning & Computer VisionEducationNazarbayev University August 2023 - May 2027BSc Computer Science Astana, KazakhstanExperienceJunior Data Analyst — Mechta.kz June 2024 - August 2025•Analyzed the performance of the main product search service, including ad-hoc analytics, monitoring, alerts, and search relevance metrics.•Worked on improvements to search index distribution and query processing logic within the search service.•Analyzed user search queries and improved product search relevance by adjusting search con figuration and keyword prioritization.•Built query filtering funnels to classify user search requests and support further search quality analysis. •Worked on automation for recognizing categories of low-frequency two-word search queries.Key Achievements:•Identi fied incorrect category prediction behavior in the search model and developed a query detection funnel; automated query labeling, which increased product clicks by 5% and conversion by 2% . •Improved keyword prioritization logic, increasing company revenue by 1.5% and improving the average product position in search results.•Updated the dictionary for the new search service, resulting in a 4% increase in product clicks and a 1.5% increase in conversion. •Analyzed consistency between multiple search services and added comparison logic between their outputs, improving search relevance and increasing NDCG by 1% . Research Assistant — Computer Vision September 2025 - December 2025•Worked on deep learning-based computer vision research, including image preprocessing, model training, evaluation, and visualization.•Prepared and processed image datasets using cleaning, normalization, augmentation, and data quality checks.•Trained and evaluated neural network models for image classi fication, segmentation, and image-based analysis.•Contributed to computer vision research involving image segmentation, synthetic image generation, and microstructure reconstruction using deep learning models.Key Achievements:•Developed a structured preprocessing and evaluation work flow for computer vision experiments. •Improved experiment reproducibility by organizing datasets, notebooks, training scripts, and evaluation outputs.•Implemented visual analysis tools to compare model predictions and identify common error patterns.•Built evaluation notebooks to compare generated and real image structures using visual metrics and statistical analysis.Projects — github.com/JokeM543 Fruit & Vegetable Image Classi fier Tech stackDevelopment: Python, Git, Jupyter Work with data: PostgreSQL, Pandas, NumPy, SciPy, Scikit-learn, PyTorch, TensorFlow, OpenCV Visualization : Matplotlib, Power BI, Tableau API development: FastAPI, SQLAlchemy Statistics: Mathematical and Statistical Analysis, A/B testing