Guli Zholmagambetova Email & Phone Number

Moscow, Moscow City, Russia

Lausanne, Switzerland

The École polytechnique fédérale de Lausanne (EPFL) is a research institute and university in Lausanne, Switzerland, specializing in the natural sciences and engineering. With over 350 laboratories and research groups on campus, EPFL is one of Europe’s most innovative and productive scientific institutions. Ranked top 3 in Europe and top 20 worldwide in many scientific rankings. - Designed and handled the AR 3D pose detector evaluation pipeline for carpentry tools (TTool).Utilized Python and NumPy to compute positional and rotational errors, involving Euclideandistance for positional error and vector-based angle differences for rotational error. Conductedstatistical analysis, including mean, median, standard deviation, and quartiles, and presentedthem in LaTeX tables and boxplots. Evaluation was performed using physical carpentry tools.- Designed End-to-End Machine Learning pipeline APIs to facilitate the addition of new classes tothe deep learning model, incorporating Postgres DB, RabbitMQ, and FastAPI services within aDocker Compose environment. Established the infrastructure by configuring databases andmessage queues on the server. - Developed a utility feature within Augmented Carpentry software utilities enabling the export of3D coordinates for both the Toolhead and Hole, specifically for use in TTool evaluation. It wasimplemented using C++ and CMake. Also, designed and integrated the user interface buttons forthis functionality using C++ and ImGui- Enhanced the Augmented Carpentry software by developing a utility feature in C++ that records UI interactions as videos and captures colored screenshots, leveraging ffmpeg, OpenCV, and OpenGL. Implemented efficient frame capture in a background thread and integrated UI controls for both video recording and screenshot capture

Astana, Kazakhstan

KazDream Technologies is one of Central Asia's biggest (TOP 5) Software Companies with +400 employees. My role has been centered on the development of Computer Vision-based software products.- Implemented an Age-Gender & Clothes Recognition features to the Video Analytics software product usingPython, ONNX-runtime, Docker-Compose, RabbitMQ- Deployed Face Detection model on Raspberry PI 3 (32-bit) using ONNX-Runtime, OpenCV, Python- Researched Cross-Compliation processes (AMD, ARM) with CMake, Toolchain for machine learning models- Developed Real-time Face Detection module for a Mobile App in Real-time Surveillance System using C++ (processing on server)- Researched & Trained models for Face Attributes Recognition using YOLO on PyTorch.- Conducted end-to-end process of building a machine learning module: starting from Data Collection, Labeling, Training, and Deploying the model- Optimized Machine Learning models using ONNX. Deployed machine learning models/APIs to production with Docker & Docker Compose, closely working with the DevOps team- Contributed several features and services to the Face Detection & Recognition software productusing Python, FastAPI, and Docker- Provided Technical Support for the Face Recognition Software Product, dealing with thesoftware's bugs using Python, MXNet, Postman

Astana, Kazakhstan

- Hold image data gathering, parsing, and image dataset preprocessing in Python, Selenium- Automated Image parsing and dataset digitization processes via Airflow, Apache Kafka thussaving time/focus for developers. It includes interactions between PostgreSQL, MongoDB, andMilvus Vector-based DB

Astana, Kazakhstan

- Developed Clothes Recognition based on Mask R-CNN algorithm on Tensorflow Object DetectionAPI- Prepared Python API using Fast API for the Clothes Recognition model

Nanjing, Jiangsu, China

SEUIC Technologies Co., Ltd. was established in 2002, continuously invests in R&D, SEUIC masters the identification technology of barcodes and vision. SEUIC keeps providing value services and solutions to more than 100,000+ domestic and foreign users.- Offered technical guidance and pre-sales support, enhancing client understanding and engagement.- Actively participated as the engineering backbone at the International CeMAT 2019 conference.- Delivered technical presentations, demonstrating the alignment of our products with client requirements across numerous discussions and meetings.- Supplied the sales team with comprehensive technical and product support materials, enabling more effective client interactions.- Provided seamless communication by translating documentation from Chinese and English into Russian.

Магистр, Control Science And Engineering, 3.85

Activities and Societies: Relevant Coursework: Stochastic Mathematics, Linear System Theory, System Modeling & Identification, Video.

Bachelor'S Degree, Major In Automation

Activities and Societies: Relevant Coursework: Calculus I, II, General Physics, Circuit Analysis (Digital, Analog), Linear Algebra.