Iranian Taxpayer System Private
NodeJs
NestJS
REST API
TypeScript
Cryptography (JWS/JWE)
Tax Compliance
Government Integration
Microservice
Developed a production-ready NestJS module for Iranian Tax Authority API integration, implementing JWS/JWE cryptographic security protocols for electronic invoice submission with automated validation against RC_IITP.IS_V7.6 specifications
Built comprehensive tax compliance solution supporting 11 invoice patterns (sales, export, contracting, securities) with TypeScript type-safety, real-time invoice validation, status inquiry APIs, and fiscal memory management for government taxpayer system integration.
n8n Node Instagram Integrations Open Source
n8n
NodeJs
npm package
TypeScript
Instagram
Community Node
Automation
Developed TypeScript-based n8n community plugin for Instagram Graph API integration with OAuth2 authentication, and webhook-based messaging (+40 GitHub stars, published on npm).
Engineered 20+ API operations including automated DM responses, comment moderation, content publishing (posts/reels/stories), and real-time event triggers with HMAC SHA256 security validation.
Bone Age Assessment for Pediatric Patients Private
Python
PyTorch
Deep Learning
Computer Vision
Medical AI
This Master's thesis project involved developing an AI system for pediatric bone age assessment using deep learning techniques.
The system analyzes X-ray images of the hand and wrist to determine the developmental age of pediatric patients,
which is crucial for diagnosing growth disorders and monitoring treatment progress.
The project utilized convolutional neural networks with a custom architecture optimized for medical image analysis.
I implemented various preprocessing techniques specific to X-ray images and conducted extensive experiments to
improve accuracy and reliability. The final system achieves results comparable to human radiologists.
AI-based Assistance Tools and Chatbots Archived
Python
FastAPI
Docker
MLOps
NLP
Developed highly scalable AI-based assistance tools and chatbots capable of handling up to 10,000 requests per second.
These systems were built with a microservices architecture and containerized using Docker for easy deployment and scaling.
The project involved creating robust API endpoints with FastAPI, implementing efficient database connections,
and setting up monitoring and logging systems. The chatbots were trained on custom datasets and optimized for
low-latency responses, making them suitable for high-traffic applications.
Custom Educational CMS Archived
Laravel
PHP
MySQL
Bootstrap
Built a powerful content management system for educational scheduling as part of my Bachelor's thesis.
This system focuses on class, student, and teacher management, providing a comprehensive solution for educational institutions.
The CMS includes features such as automated scheduling, conflict resolution, attendance tracking,
grade management, and reporting tools. The system was designed with a user-friendly interface to accommodate
users with varying levels of technical expertise.