Sarp İpekçi Email & Phone Number

Istanbul, Turkey

I am currently a Software Developer on the Anti-Money Laundering (AML) team at Fineksus. My work focuses on developing innovative solutions that enable financial institutions to securely handle transactions like SWIFT, Wire Transfers, and EFTs, while complying with AML regulations. Products I work on address key challenges in the financial sector, such as monitoring customer transactions and behavior, conducting risk assessments, and implementing blacklist filtering. Projects I work on concentrate on tracking customer and transaction activities, managing risk and compliance, thus helping institutions protect their integrity and security. Additionally, I contribute to solving industry challenges by enhancing financial transaction and customer interaction management through KYC protocols and other processes.

Amur Oblastı, Rusya

Moscow-based ESTA Construction Company's employers are Sibur and Nipigas, located in Svobodny, Amur. In the technical office of the construction site of the platform that will supply natural gas to China. The technical office requirements analysis, planning, and programming of the integration process. After the requirements, analysis is complete by using NetFramework, and the MSSQL server Construction Management Software program is built. Excel data is included in the construction progress report. Export to Excel with the Management Software program that created the processing part has removed, and data entered through the program's website, construction progress report is created automatically by the system. The program allowed instant monitoring by authorized users. Data MSSQL It is recorded in the server. Users can download the data of the construction progress report in Excel format and be able to upload it to the system. In addition to speeding up the program, it also provided instant viewing.

İzmir, Turkey

Mell Beauty Center located in the Torbalı district of Izmir, provides customer registration, stock registration, and customer receipts. They were trying to manage their services, session, and installment information via Excel. Using Winform beauty center automation built. As requested in the program, users enter the session and installment information in response to the service they receive. Ease of use was ensured by automatically dividing it into months and amounts. For monthly and daily appointments, a system was created for service satisfaction and follow-up by giving warnings to users. Embedded database efficiency profited from storage and server cost. Users can additionally provide customer, service, session, installment, debt, appointment, and stock information can be printed in Excel, PDF, and Word or in Developed environments where they can restore the system.

Ankara, Turkey

The purpose of the development of the project is to distinguish between people and objects by using image processing and artificial intelligence and classifying objects (car, plate, seat) by dividing them into subheadings. Program developed in the Python environment. A dataset was creat to train the program. In addition, to the photographs of objects and people, videos are included in the dataset. With the VideoCap feature, every frame is converted into a photo, increasing the target probability for the train. In the training process, CPU and GPU machine learning is performed using Tensorflow and OpenCV image processing technologies. With additional learnings, the target probability increased, and the results from the learning record were in the folder for the follow-up project (Auto Pilot System). In this way, We created the first training part of the continuation project. Then the system was put to the test.

Nicosia, Cyprus

The autopilot system is the second phase of the image object classification project. Autopilot project is the project of following the vehicle's lane, sensing the road surface, recognizing other cars, and reacting to environmental factors. The software was prepared in Python programming languages and transferred to Raspberry Pi 3 b+ environment. Instant analysis and imaging processes were performed by utilizing the processor of Raspberry Pi 3 b+. For image processing, a stereo camera is mounted on the Raspberry Pi 3 b+ and calibrated. Calibrations were made using image processing technology in the Python environment. A dataset folder has been created for the autopilot program to analyze. The dataset of the autopilot program used lane photos, different vehicle photos, and road surface photos. In the object classification project with image processing, the data set created in the autopilot program was added to the human, environment, and object classes. In artificial intelligence training for autopilot programs, the environment is prepared. CPU to increase the training speed of autopilot software and the GPU trained, appropriate CNN and DNN neural networks and data structures created. With KNN and RANSAC algorithms, reaching the target and increasing the success rate have been achieved

Ankara, Türkiye

I worked with expert staff during my internship in the Technology Department of DSI, a state institution. I took part in the camp project. The camp project is a system that allows DSI employees to spend their summer vacation at the DSI holiday camps they choose for their summer vacation, using their points and preferences. Within the Yetki.Net system used in DSI, it started with the scoring part of the camp project software found. Employees before camp preferences entered the scoring category. Points consist of the employee's start date, the number of days worked, positions, and regions they worked. The scores are stored in the Oracle database, and the first stage was created by obtaining employee information from there. The front of the program is with MVC, and the backend programming is built with C#. A new selection page has been created in Yetki.Net for the summer camp selection. Employees who prefer more than one camp are placed in order of preference according to their scores if they exceed the main points given to the camps. I created a reserve list for the most preferred DSI camps. After any cancellation, the person on the reserve list is placed by the system according to the points.

Ankara, Türkiye

Connected2.me is social media platform like Instagram and Twitter. Connected2.me allows you to match another person and chat with them by your username or anonymously. During my internship period, I took part in the database team. I developed a program to detect users violating the rules and take necessary criminal action. OCR program that uses Python programming to quickly identify violators, make quick decisions, and enforce them instantly is built. The OCR program automatically detected the user's swearing, insulting and racist expressions and matched them with processed insults, profanity, and racist words found on the AWS server. Users who use the application if the insults they type and say match the swear words found on the AWS server, penalties are applied. OCR detection of public stories or videos phase passed. A local website was created using Python, and stories that seemed to be problematic via e-mail were taken as URLs and divided into frames with the VideoCap -OpenCV program prepared in Python. It has been filtered in the OCR program. If there is a problem with the user's stories, they are penalized. MySQL database has been created for penalty registration transactions. The identity information of each user is penalized using the database stored in the database. The data of users who violated the rules stored in the database was backed up by sending them to the prime server.