I have a strong foundation in building scalable systems with Python, C#, Kafka, Flink, and cloud technologies. I am passionate about solving complex problems through clean, efficient code and integrating AI & Computer Vision components into production-grade pipelines.
An end-to-end Multiple Object Tracking (MOT) system designed for retail customer flow analysis.
- Engineered using YOLOX and Detectron2 with a modular, polymorphic architecture.
- Integrated ByteTrack and BoTSort for trajectory consistency and path fragmentation handling.
- Optimized the pipeline to achieve full real-time processing (1:1 ratio) for 25 FPS video streams.
Developed under Vast Data & KamaTech Bootcamp A high-throughput microservices system designed for processing real-time agricultural IoT and image data.
- Implemented a robust data pipeline: Image uploads via MQTT β‘οΈ stored in MinIO (Local S3) β‘οΈ event streaming via Kafka β‘οΈ low-latency ingestion and windowing in Apache Flink with a Dead Letter Queue (DLQ).
- Developed a soil moisture prediction microservice wrapping a Deep Learning model optimized with ONNX.
- Built a desktop application GUI using PyQt6 for displaying maps, irrigation status, and data history.
- Designed Protobuf/JSON schemas with automated validation, CI/CD via GitHub Actions, and raw data cleaning in Apache Spark.
A modular, high-performance Backend system built with modern .NET enterprise architecture patterns.
- Designed an layered architecture using ASP.NET Core Web API and Entity Framework Core connected to a SQL Server database.
- Implemented the Repository Pattern and Dependency Injection alongside optimized LINQ queries.
- Developed custom Stored Procedures, Views, and Triggers for reliable business logic processing and secure Authentication/Authorization mechanisms.
A full-featured responsive SaaS platform for managing enrichment programs and clubs, hosting three custom user roles (admin, teacher, student).
- Developed a responsive Frontend with React, Redux, and PrimeReact.
- Built a secure REST API with Node.js, Express, and JWT-based Role-Based Access Control (RBAC).
- Designed and optimized a dynamic database schema in MongoDB using Mongoose to handle attendance tracking and payments.
β‘ Most Used Languages:
- π§ Email: tamar45334@gmail.com
- πΌ Resume: Available upon request