Bio

I’m a Senior Datawarehouse Analyst with Data Analytics Master Degree. Currently working as a Senior Datawarehouse Analyst in Turkcell at AI and Business Intelligence team. I have experience as a data analytics projects business owner and as a sales analytics specialist. Also have experience in technology retail for 5 years, 1.5 years of it as planning team head.

I completed my master’s degree in Data Analytics at Sabanci University in 2020 and also wrote data analytics content for the Analytics Vidyha Medium blog page and H20.ai. You can see reach these articles on my portfolio website.

Portfolio: http://www.semihdesticioglu.com

Medium Articles: https://semihdesticioglu.medium.com/

H2O.ai Articles: https://h2o.ai/blog/how-to-create-your-spotify-eda-app-with-h2o-wave/

GitHub: https://github.com/semihdesticioglu

Skills:

Programming: Python (Pandas, Sci-kit-learn, PySpark, Kafka Python Driver), SQL, PL/SQL, NoSQL
Data Engineering: Data Modeling, Database Design & Management, ETL, Hadoop HDFS, Hadoop YARN, Hadoop MapReduce, Hive, Sqoop, Kafka, Delta Lake, Airflow, Spark
Data Visualization: Matplotlib, Microstrategy, Tableau
Databases: Oracle, BigQuery, Firebase, Elasticsearch, Cassandra, MongoDB
Cloud Services: AWS, Google Cloud, Heroku, IBM Cloud, H20.ai
Web Scraping: Selenium, Beautiful Soup, Scrapy
Containerization: Docker, Kubernetes
Linux Basic and Bash Scripting
Web/App Development: Flask, HTML, CSS, FlutterFlow
Modelling: Supervised & Unsupervised Learning, Time Series Forecasting, Neural Networks, Deep Learning

Portfolio

Real-time IoT Sensor Data Dashboard

This project aims to visualize sensor data in real time by utilizing Spark Streaming, Kafka, and Kibana. The dataset contains data from 255 sensors in 51 rooms across 4 floors of the Sutardja Dai Hall at UC Berkeley. The data includes CO2 concentration, humidity, temperature, light, and PIR motion sensor readings. The project involves

preprocessing the dataset with PySpark,
producing it into a Kafka topic,
consuming the topic with Spark Streaming,
and finally visualize the data with Kibana on ElasticSearch.

Github Repo : https://github.com/semihdesticioglu/Real-time-IoT-Sensor-Data-Dashboard

Spotify EDA App with H2O.ai

I wrote an article about my Spotify Data Analysis Dashboard project which I produced with H20.ai tools. With this tool, you can upload your stream history and see your music history analysis.

Article Link: https://h2o.ai/blog/how-to-create-your-spotify-eda-app-with-h2o-wave Github Repo: https://github.com/semihdesticioglu/SpotifyWaveApp

Covid-19 Dashboard Turkey 💻 📈 by Using Python 🐍

End-to-end dashboard project to visualize COVID-19 data of Turkey. It is built using Python’s Flask Web Framework and deployed on Heroku cloud. Plotly library from Python is used to visualize data. For front-end development bootstrap, javascript, and CSS are used together on HTML.

Link to the app: https://covid19turkiye.herokuapp.com/

Turkcell AI Vision Challenge Contest / Deep Learning Project

Turkcell AI Vision Challenge competition was held in my company internally and I won the prize with my winning project. Out of 120 projects, it is awarded as 1st. I created a CNN deep learning algorithm using Pytorch library and pre-trained Resnet-18 model with the transfer learning approach.

Github Repo: https://github.com/semihdesticioglu/t-ai-vision-challenge

Airbnb Istanbul Data Analysis

I analyzed Airbnb Istanbul data and answered some questions about the data. The project’s scope is exploratory data analysis with visuals and maps, applying a statistical test to data to answer the questions correctly.

All analyses can be found on Medium Post : https://medium.com/analytics-vidhya/airbnb-istanbul-data-analysis-b510c98a37e4

Github Repo : https://github.com/semihdesticioglu/airbnb_istanbul

Amazon & MediaMarkt Smartphone Prices Web Scraping Project

This project aims to get smartphone and other devices’ current prices from Amazon and MediaMarkt websites by web scraping. Selenium, BeatifulSoup, and Request libraries are used for scraping. The scraping algorithm collects product models, brands, and old and new prices from Amazon & MediaMarkt websites and builds Pandas dataframes with these collections.

I was nominated to CXO prize with this RPA tool. It is used for monitoring the competitor prices in Turkcell. Tool was sending daily auto-emails with scraped price lists.

Github Repo : https://github.com/semihdesticioglu/WebScrapingProject

Music App Churn Prediction with Pyspark on IBM Cloud Pak

In this project, I tried to create a churn model with the given data. Creating new features in the feature engineering section added valuable improvements to the model. Data processing is done with PySpark in distributed environment.

You can find this project’s Medium article on this link : https://medium.com/@semihdesticioglu/predicting-music-app-churns-with-machine-learning-sparkify-project-dfc2ac2c695c

Disaster Response Project - Multioutput NLP Classifier

This project aims to build a Natural Language Processing (NLP) model to categorize messages for possible disaster events that could happen in the future.

In this project, I have built a Multioutput Classification model to predict message categories in future possible disasters. There are 36 pre-defined categories, and examples of these categories include Aid Related, Medical Help, Food, Death etc. By classifying these messages, we can help the government or other aid organizations prioritize critical messages in possible disaster events.

Github Repo : https://github.com/semihdesticioglu/twitter-nlp-classifier