Uğur Civelek Email & Phone Number

İstanbul, Türkiye

İstanbul, Türkiye

I contributed to many analytical and AI-based projects and have worked on various aspects and phases of them. The two most significant projects I contributed to are the following:- Obase Smartfind Search EngineDesigned to operate within e-commerce websites, Smartfind seeks to get the best possible matching and ranking order among the products sold in a website based on the user query, using various machine learning and NLP techniques embedded into it. It's a multi-cloud SaaS solution and currently able to run on AWS and Azure. The following features are incorporated into this solution:* Typo tolerance* Autocomplete* Morphological analysis* Category prediction* Semantic similarity analysis* Algorithmic ranking* Analytics and insightsTechnologies: Python Ecosystem, Docker, Git/Gitlab, AWS Cloud Technologies- Yıldız Holding CDP ImplementationUsing Microsoft’s CDP solution, Customer Insights, we integrated and merged customer-related activity data sources from various subsidiaries of one of the largest holdings of Turkey, Yildiz, and conducted many analytical operations on CI platform according to the needs for understanding customer behaviour. Besides rule-based segmentation and measuring techniques, we also made use of platform’s built-in machine learning pipelines for churn prediction, product recommendation and customer lifetime value(CLV), providing Yildiz a more comprehensive and complex analysis to understand their customers’ behavioural patterns. Following tasks are accomplished:* Match and merge customer data coming from different data sources* Customer segmentation* Desigining measures* Running analytical models* Analytical insights* Exporting analytical items to social media platforms for further campaign operationsTechnologies: Microsoft Customer Insights, Azure Data Lake, Azure Synapse, SQL

Istanbul, Turkey

I made assistance to a medical research project funded by TUBITAK (The Scientific and Technological Research Council of Turkey) within the body of Bahçeşehir University's School of Natural and Applied Sciences, titled "Combining the Cortical Motor and Sensory System Measurements in Parkinson's Disease Examinations with Machine Learning Methods". In particular, I focused on "deep learning" aspect of the project. In this project, - Our aim was to predict the intensity of the patients’ disease in terms of a numerical value lying within the universally-accepted measurement scale for Parkinson’s disease called UPDRS. - We used audio recordings collected from Parkinson’s disease patients at Cerrahpaşa Faculty of Medicine’s Neurology Department as our sample data. - We investigated optimal deep learning models by training and validating different architectures in the light of audio data at hand. Softwares and platforms used: R, RStudio, Keras, Tensorflow, Amazon Web Services (AWS) Technology keywords: machine learning, deep learning, image recognition, voice analysis

Master Of Science (Msc), Bilgisayar Mühendisliği

Bachelor Of Science (Bs), Elektrik Ve Elektronik Mühendisliği

Activities and Societies: IEEE Engineering and Technology Club