Bachelor of Science
GPA:2.5/4
Architecting new solutions from the ground up
Working closely with the quality assurance teams and designers
Extend existing software solutions and infrastructure
● C# ● CosmosDb ● Azure Webjobs ● Azure Storage ● MsSQL ● F# ● Angular ● Docker ● Azure Functions ● NancyFx ● TeamCity ● KubernetesLead a team of 5
Software design and development for a web based ERP tool
Modified existing Web Forms project to fix errors and improve Performance
Researched, designed, and implemented scalable new .NET Core project
Wrote maintainable and extensible code in a team enviroment
Implemented a unified authentication project that is used for all products
● .Net Core ● C# ● ASP.NET Webforms ● ASP.NET Web API ● ASP.NET MVC ● Angular 5 ● MongoDB ● Redis ● MsSQL ● Scrum ● MsSQL ● TSQL ● OAuth2 ● TFS ● EF Core ● TDDProvided leadership and mentoring to a team of 7 developers.
Designed and developed a real time messaging tool that is used in B2B projects.
Managed complete development lifecycle of several small-scaled projects.
● SignalR ● LESS ● ASP.NET ● MVC ● Redis ● MsSQL ● AzureWrote a data mining library
Implemented ANN, Genetic Alg., Decision Tree L. and several Clustering Alg.
Designed and developed Pinnera Analyzer Windows Form Project.
Rapidly prototyped new capabilities to confirm feasibility
Developed a web based data mining tool for SMEs
Designed and developed a web ERP tool for factories
● C# ● WinForms ● ASP.NET MVC ● ASP.NET Web API ● EF ● MsSQL ● MLDTSharp is a .net decision tree learning library. It uses a specialized algorithm like ID3 and C4.5 to generate a decision tree. It has the capabilities of pruning (depth,data count,output probability) and using several split qualifiers (like GiniImpurity,InformationGain,Entropy), discrete and continious features, Iqueryble data sources and using Entity Framework as a data source (Transfers processing load to db).
In my graduation project of my bachelor degree in Electronics Engineering in Istanbul Technical University, i've focused on Pattern Classification using Decision tree learning and SOM Networks. To classify different patterns like Brodatz samples and ultrasound images, I used several features of image processing. I trained a decision tree and a SOM Network, then i compared their results to see if these features are suitable for this purpose or not. In conclution, DT has a average success rate of 88% while SOM Network has %76.
Clocked is a Tile based sandbox strategy game developed with XNA 4.0. In the game, players have to collect ores with their tank fleet and use it to improve their tanks. The goal of the game is reaching the center of map which is created randomly with procedural world generation algorithms.
● C# ● XNA 4.0 ● Socket