My Projects
Background module development for corporate company
Real-time data processing with Apache Spark, Redis integration, and optimized data pipelines for enterprise-scale operations.
Mobile image processing on conveyor belt
IoT-based solution combining Raspberry Pi, servo motors, and TensorFlow for automated object detection on conveyor systems.
PetClinic Application
Full-stack veterinary clinic management system with user records, appointment scheduling, and enhanced UX design.
Audroad – FinTech AI Pipeline
Real-time audio processing pipeline built on Apache Kafka KRaft, Docker-based microservices, and AI-powered transcription with faster-whisper.
Background module development for corporate company
This application continuously monitors directories using Apache Spark and enriches the files with data retrieved from Redis. With real-time data processing capabilities, it instantly detects file changes, collects data, and integrates it with information obtained from Redis. The enriched data is then securely and efficiently transferred to various message queues and big data platforms through optimized data pipelines. This solution automates data integration and distribution processes, enhancing operational efficiency.
Mobile image processing on conveyor belt
This project provided communication of Raspberry Pi-Servo motor-Conveyor band devices using IoT fundamentals. It determined the decision structure with a customized model on Tensorflow. An open server (HiveMQ) was turned into a client for the conveyor band to communicate.
PetClinic Application
I developed a website to cater to the needs of a small-scale veterinary clinic, managing user records and addressing their requirements. The website focused on visual representation and enhancing user experience, with implementation carried out using Thymeleaf.
Audroad – FinTech AI Pipeline
Audroad is a FinTech-oriented real-time audio intelligence platform. It captures live audio streams via WebSocket, routes them through Apache Kafka (KRaft mode — no ZooKeeper) for resilient, low-latency messaging, and processes them with a Python-based AI transcription service powered by faster-whisper. Object storage is handled by MinIO, and the entire stack runs as Docker-composed microservices. The architecture is designed to handle millions of messages on minimal server resources.