Omer Aynur

OMER AYNUR Data Scientist/Machine Learning Engineer Ankara Turkey -- Data Scientist/Machine Learning Engineer with a demonstrated track record of excellence in research and development, spanning over five years. Background includes Computer Vision, NLP, Reinforcement Learning, Time Series Forecasting, Medical Imaging, and Architecting end-to-end Machine Learning Systems . Throughout my career, I have consistently delivered exceptional results, developing, deploying, and monitoring advanced Machine and Deep Learning models that have been instrumental in building successful customer applications. Skills Python, Node.js, FastAPI, Flask ,Express AWS,EC2,EKS,S3,Docker,Kubernetes,Jenkins,Rancher,Git,Linux Pytorch, Tensorflow, Keras, Hugging Face, Sklearn, XGBoost,LightGBM, CatBoost, NumPy, pandas, boto3, pyspark Kubeflow ,KServe, BentoML, Seldon, MLFlow, Airflow MongoDB, Kafka,PostgreSQL,Redis, Spark, Databricks, RabbitMQ Computer Vision, Predictive Modelling, Medical Imaging, NLP, Time Series Forecasting, MLOps, Machine Learning Systems Work History 2023-08 - Current Senior Data Scientist/ Machine Learning Engineer Vodafone Netherlands, Remote Implementing Recommender Systems for Vodafone products and test cases of Q&A Team. Prototype, Develop, and Optimize LLM based applications using AzureOpenAI and ChatGPT for Knowledge Based Conversational Chat Bot and Keyword Extraction. Utilizing LLMs to extract keyword from resume of candidates for job matching for Vodafone HR team Heavily using Langchain, LLaMaIndex, for creating LLM applications and ChromaDB, Pinecone for VectorDatabase 2022-04 – 2023-08 Senior Data Scientist/ Machine Learning Engineer Presify, Ankara Implementing Time Series Forecasting models to predict the power generation of thousands of renewable plants such as Solar and Wind for both intra-day and day-ahead markets using LSTM, Transformers , XGBoost, LightGBM, CatBoost Random Forest. Prototype, Develop, and Optimize Time Series Forecasting models to predict Electricity Demand for various regions of Turkey using LSTM, XGBoost, LightGBM, and Random Forest. The work in this problem involves collecting, cleaning, and preprocessing large volumes of data, using advanced techniques to remove outliers and handle missing values. Implementing Anomaly Detection algorithms for the ENTSO-E database ( European Network of Transmission System Operators for Electricity ) including all European Countries' electricity dataset which consists of load, generation, transmission, and balancing domain. Used LSTM, XGBoost, CatBoost, and RandomForest. Optimized models' hyperparameters and thresholds to achieve high accuracy and low false positive rates, while accounting for the imbalanced nature of the data. Conducting extensive data preprocessing and feature engineering to improve models' accuracy and scalability. Designing Machine Learning development and production environment with microservice architecture in Kubernetes clusters and Kubeflow. Deploying machine learning models into production using various deployment strategies, such as REST APIs, KServe, BentoML, and containerization technologies like Docker and Kubernetes. Continuously monitoring and training deployed machine learning models using performance metrics and user feedback, and made adjustments and improvements as necessary to meet business requirements. Developing and maintaining core microservices for production environment using FastAPI, Flask, Node.js, Kafka, Redis, PostgreSQL,MongoDB in Kubernetes cluster. 2020-04 - 2022-04 Data Scientist/Machine Learning Engineer Titra Technology, Ankara Processing, and analyzing high dimensional 3D TOF MRA images under supervision of radiology group. Develop, optimize and implement 3D Vision Transformer, 3D Res U-Net, and 3D Attention U-Net based Deep models and their variants to provide diagnostic assistance for intracranial pathologies like aneurysms and ischemic strokes using Tensorflow, Pytorch, and Sklearn. Deployed production-grade models to tens of hospitals, optimizing their diagnostic and treatment capabilities. Reduced workload of radiologists ,Increased diagnostic accuracy of radiologists in unruptured aneurysms by %10 with assistance of developed models. Designing and implementing a Reinforcement Learning-based control system for the take-off process of UAVs , using DDPG as the underlying algorithm. Used OpenAI Gym simulation environment to train and test the control system. Successfully developed and deployed Object Detection and Segmentation models for embedded devices such as UAVs, resulting in increased accuracy and efficiency in aerial surveillance and monitoring operations. Developed custom CI/CD pipelines and workflows that improved the deployment process of machine learning models, leading to a significant reduction in timetoproduction using Jenkins and Kubeflow. NLP Engineer/Data Scientist 2018-06 - 2019-10 Etiya Technology, Istanbul Used Machine Learning and Statistical Learning techniques to develop robust models for large-scale Turkish and English sentences to tackle complex business problems like sentiment analysis, abuse detection, chunk extraction. Extensively used RNN, LSTM to model and analyze the sequential behavior of customers to extract insightful patterns and trends from large and complex datasets to provide meaningful and actionable insights for businesses. Developed, Deployed, and Monitored state-of-the-art sequential Recommender Systems for TV contents of Videotron Telecom in Canada which has over 1 million customers. Extensively Used collaborative filtering, content-based filtering techniques. Deployed Machine Learning models on AWS EC2 instances using Docker , ensuring high availability, scalability, and reliability of the models. Education 2013-09 - 2018-05 Bachelor of Science: Computer Engineering Erciyes University GPA: 3.59/4.0, Graduated with Highest Honors in the Department