George Zervas

Experience

• 

  University of the Aegean

  Sept 2013 - Oct 2018

  As a member of the Aegean Robotic Team, we created a humanoid. His name is El Greco and it consists of two arms, a head and a body. It is able to move in any direction using tracks. The ability to make gestures, recognize the environment and people around it is also implemented. It can speak Greek and receive orders via voice command. The whole robot is 3D-printed and it has servo motors inside his body to move all the moving parts. We use python to program the robot. The user of the robot can easily handle it through the GUI that we have created. This GUI consists of three different types of functions, which are low-level, middle and high-level functions. The robot has a camera on his head and El Greco has the ability to recognize text in almost any language and read it. The face recognition was made with OpenCV techniques and image processing algorithms. Show less This is my diploma thesis. I made a data analysis of gas retail data. I started with Problem Framing and Data Understanding, to take a first view of the dataset. Then I use Excel to clean and transform some data types of them. I also do some changes in Python Jupyter, In Model Evaluation I chose to create a custom Linear Regression, combined with non-consecutive seasonality terms (weeks) and as well as using the rest available variables. I made also an Autocorrelation Analysis which reveals us the critical weeks of the dataset correlated with a delayed copy of this set (lags). I used all these discrete events (weeks), distinguishing independent (X) and independent (Y) values and I trained my model. Finally, I predict sales with a score of close to 92%. The evaluation of the model includes the evolution of the analysis (from the use of a few data to the exploitation of all) and of course the calculation of metrics STE and MSE. Show less

  • El Greco Humanoid

    Feb 2017 - Oct 2018

  • Study and Implementation of a Predictive Model Of Forecasting Sales In Fuel Retail Data

    Apr 2018 - Sept 2018

  • Graduate Student

    Sept 2013 - Sept 2018
• 

  ITU - Information Technology University

  Apr 2015 - Apr 2015

  ITURO 2015 Robot Olympic Games

  We create an autonomus robot that can move an a track and detecting and extiguishes the fire. We use Arduino UNO and several Ultrasonic sensors to achive wall follow. To detect fire we used specific sensors. We also put a blower at the top of the robot to blow the candles that was in the track.

• 

  1st Crowdhackathon-SmartCity

  May 2017 - May 2017

  CitySens

  We took part as an IT Crowd team in the first Crowd Hacathon contest in Athens. We created a chat bot on the facebook messenger, which involves sending complaints and sending ratings for certain municipality services. It's name is CitySens and it is available on Facebook. We created also a Website which support this bot. This bot creates open data which, through appropriate processing techniques, can provide useful information to the city's administrative staff. Identify weaknesses and tackle them effectively. We used Node-JS to create the bot and CSS-HTML5 for the Website. Show less

• 

  PwC Greece

  Jun 2017 - Sept 2017

  Intern in Business Intelligence and Analytics Department

  I learned the axes on which BI is based and I fully understood the concepts and tools I used. I worked with Targit and PowerBI, which are visualization programs and helped the BI team to visualize the results of data analysis. Then I studied architecture and how a BI system is being implemented. In more detail, I set up all the necessary tools/programs that include a complete BI system ( ETL Tools, SQL Server Integration Services, Data Warehouse, SQL Server Analysis Services, Visualization). I did the complete process followed by a BI team to get the data, configure it properly, set up the Data Warehouse and finally present it using the right tools (Reporting and Analysis). Show less

• 

  CERN

  Jul 2017 - Jul 2017

  Cern

  As part of an educational trip organized by IEEE at the University of the Aegean, we went to CERN. Here we went to the Cern sites we saw the ATLAS control centre we saw their Datacenter as we travelled to the place where the tests for the durability and maintenance of the accelerator components are done. Finally, some lectures by physicists working on Cern gave us a sample of their work and the evolution of the experiments there. It was an unbelievable experience.

• 

  SAP

  May 2018 - Jun 2018

  Intern in Data Science Hub

  I worked in Data Science Hub Team. My responsibilities were to prepare the Data, clean them, update them, bring them in an appropriate format. I was part of all the ETL process from Raw Data that we were given to HANA Import process. Then with my colleagues, we set up all the necessary tables, calculation views and Views to organize the data for the Analysis. In the Analysis section, we used the Predictive Analytics Algorithm (Multiple Linear Regression combined them with Business Indicators) to find all the Data insights. The results of this action (Analysis and Report) was visualized, via UI5. The programs that I used in SAP was SAP HANA Studio, Predictive Analytics and the BI tool Lumira. Show less

• 

  Hellenic Army

  Jan 2019 - Aug 2019

  Project Manager, Full Stack Developer, IT Support

  During my military service, I was assigned by the Army General Staff to manage and design a new information system that will serve the new and higher requirements than those of the existing system. This information system is created to support, manage and organize military equipment to optimize the performance.Ι worked as Project Manager as well as Full Stack Developer ( front and back-end ) of this project, I made the maintenance and the administration of the existing system and of course, I provide IT support to any kind of problem that appeared. The responsibility for completing the new project was great, but I liked to work under pressure and the burden of responsibility. Show less

• 

  Aegean Airlines

  Oct 2019 - Jun 2023

  • Crew Forecasting & Man Power Management : predict the future crew demands to cover company’s operational needs and measure the possible impact of different business strategies.• Crew’s Monthly Flight Schedule : Prepare and develop all the necessary components, files analysis and checks to provide insights and support the production of Crew Flight Schedule.• Bi Reports (Reporting & Data Engineering) : Design & Develop Bi Reports to support Business decisions and Strategies about Crew Fatigue Minimization, Rest Hours, Flight Hours, and Leaves. Use Bi to prepare files about system uploaders to avoid large scale manual actions. Show less Main Responsibilities:• Understand business requirements and design data models to convert raw data into meaningful insights.• Process & Production Flows - Optimization & Automation / Standardization• Design, develop, and deploy Reports (Business Objects, Power Bi, SSRS) and perform efficient detailed analysis.• SSIS Projects Development• Digital Transformation of the Finance Department• ETL Process through Python Scripts from several raw data sources.• Responsible for Finance Dep. File Archiving Automated Process (Python Script)• Enhancement of Sales Data Integrity with validation tests and developed automated processes • Ad-hoc Reports & Analysis for C-Level• Platform Development about Employee Productivity Score (Linear Regression Model) Show less

  • Senior Business Data Analyst / Crew Planning

    Jun 2022 - Jun 2023

  • Reporting Data Analyst / Finance

    Oct 2019 - Jun 2022
• 

  OAED

  Dec 2021 - Dec 2021

  HireHack

  HireHack team won the 1st price in the 1st "be a part of hack OAED crowdhackathon". We developed an engine which has the ability to read, classify and match employee's resumes with company job descriptions. This engine acts as an assistant of human resources agents and works for them during the first stages of job finding. This procedure until now is manual in OAED Institution and this solution gives the opportunity for less time-consuming action and efficient results. The classification process and the matching next are key points of the engine. The code is written in python and the classification of the resume in the appropriate category is based on keywords that exist on the internet.Phase II of this engine is to act as a crawler, collect keywords from recruitment sites and then do a better classification because of the brighter knowledge. Show less

• 

  Alpha Bank

  Jul 2023 - now

  Senior Data Engineer