Projects
Microservice for Monitoring Tool Analysis (Flask, Python, Prometheus, Grafana, Zabbix, Dynatrace) [2024]: Microservice developed using Python Flask web framework mainly to monitor the HTTP request frequency and availability of the service runtime by different monitoring tools like Prometheus, Grafana, Zabbix, Dynatrace. The main objective was to gather data for a comparative analysis about the overheads (e.g. CPU and Memory usage) of these monitoring tools while monitoring a microservice.
Detecting COVID-19 from Chest X-Ray Images (Python, TensorFlow, Scikit-learn) [2023]: Comparative classification of three classes of chest x-ray image dataset (COVID-19, Normal, Viral Pneumonia) to detect COVID-19 using different models like CNN, K-NN, SVM, FFNN and Ensemble Learning.
IUT 10th ICT Fest Website (Laravel, PHP, JavaScript, MySQL) [2019]: Registration and payment platform for an event deployed with usage from about 300 participants of IUT 10th ICT Fest 2019. It includes registration for 11 events and has an automated confirmation email feature sending the participants a secret key which was used in payment procedures.
OpenShop (PHP, MySQL) [2018]: E-Commerce website with features like cart, wishlist, order, and shipping management.
Parking Management System (Java, Oracle SQL) [2017]: The parking system was built using JAVA FX having features to input different categories of vehicles and allocate a parking space at particular location based on their vehicle class.
Checkers Game (C++) [2016]: It is a classic checkers game with a 30 seconds timer for each player to give a move. If a player fails to give a move then a random valid move will be given automatically after 30 seconds.
Desktop applications for a physiotherapy clinic (C#, WinForms .NET, SQL): Multiple Windows desktop applications each with their own features together handled the system of a physiotherapy clinic. The old applications of the clinic were developed in Visual Basic 6. Keeping the functionality of the legacy code similar with added new features, the new desktop applications were developed using C# WinForms .NET framework v4.7.2 with Access DB and SQL Server as the database with Three-tier Architecture following SOLID principles. TrueDBGrid was used to show the tables on the forms. DevExpress was also used to develop some interface of the applications. The overall performance of the .NET applications were improved by about 25% than the Visual Basic application using Caching Mechanism with some start-time tradeoffs.
Therapy Application: It was the main application to control the overall system of the physiotherapy clinic. It showed the details of patients, doctors and insurers of the clinic. Each patient was a part of a series which had its treatment, prescription, appointments, doctors, insurers, invoices, patient history etc. It had the information of the ongoing or future treatments of a patient by a certain doctor and the required prescriptions for the treatment. It could calculate and generate an invoice of the treatment based on the insurance of the patient. From this application we could open the Appointment application, Roadmap application, Fitness application and Settings application.
Appointment Application: The application contained a calendar showing the appointment slots of the patients and colored section of the availability of each doctors. The calendar was fully customizable. An appointment could be created, edited and deleted from the calendar. Each appointment also had a colored band denoting the type of appointment like massage, acupuncture etc. It could also make appointments for Fitness application. Those appointments could be linked with the treatments from the Therapy application. Each treatment of the patient can consist of multiple appointments. When an appointment was completed it would notify the therapy application and that application would generate an invoice.
Roadmap Application: The application contained customizable patient profile including its history and ongoing conditions. Each patient profile could be made with a customizable template as the doctor wants it to be organized. Some templates could contain images with labels along with a table for notes and histories, and also other templates may contains just rows of data. Drawing was also possible above the image to denote some specific markups. The template were customizable by selecting different tools like graphic image, image with tables, dynamic rows and columns, text box, blank space etc. This templates of the patient profile were accessible from the Therapy application.
Fitness Application: The application contained the fitness gym profile of a patient of the clinic. It would hold the information of the subscription of their gym package. Based on their subscription it would generate an invoice monthly or annually. It would also keep record of their prescribed fitness training and their gym start datetime and stop datetime.
Settings Application: The application controlled the general settings for all of the applications of the clinic. For example, the text font and size of all the applications, the printer settings, the clinic opening and closing hours etc.
Generic Kiosk Web Service for a Restaurant (ASP.NET Core, MongoDB): The web service handled the food ordering of a restaurant from a kiosk. It was developed using ASP.NET Core v3.1 and MongoDB as the database. The code architecture followed CQRS pattern and Mediator pattern. Fluent validation was used to write validation codes and xUnit.net was used to write unit tests for each API endpoints. The web service had multiple endpoint REST APIs for creating customer account or admin account, to place order and payment by customers, to add or edit food items by admin etc. The orders were dispatched to the system in the kitchen by web socket. Multiple order details were kept in a message queue by RabbitMQ and they were processed and updated in the database as order logs. I was mainly involved in developing multiple REST API endpoints of this service. Jira was used as the team management tool to maintain agile methodologies. Bitbucket was used to keep the code repository and it was maintained using Git. The service was deployed in Azure with CI/CD pipeline. Increased food sales by up to 30% with upselling features and minimized order wait time.