Rez Eghbali

• Contributed to a role-based access control (RBAC) file system, enabling secure upload, download, and sharing of files with user permissions. • Refactored services towards a Clean Architecture approach, improving maintainability and reducing request latency by ~20%. • Containerized services using Podman and built RPM packages to streamline deployment and integration with Linux environments.

• Designed and implemented REST APIs with OAuth2 authentication, achieving less than 100ms latency. • Implemented idempotency in trading operations to ensure reliability and prevent duplicate transactions. • Implemented concurrent WebSocket clients with Goroutines to collect real-time market data from 20+ crypto exchanges, covering 10K+ tokens and processing 500K+ price updates per second.

• Built a distributed crawler in Golang for gathering real-time data from news websites and social media, scaling to 100K+ posts daily. • Implemented IP rate limiting and smart rotation strategy, reducing proxy provider costs by 30% while maintaining 99.9% uptime. • Designed fault-tolerant modules to handle unstable third-party APIs, reducing crawler failures by 25%.

Our team aims to find solutions to everyday problems and the usage of AI techniques to bring convenience to human lives. Currently, we focus our target on the application of deep learning in NLP scope.