Daniah Qwareq

AI/ML & LLMs
• Explored Generative AI with Large Language Models (LLMs) concepts using Hugging Face Transformers and PyTorch, gaining hands-on experience with transformer-based text generation, prompt engineering, instruction fine-tuning, and Parameter-Efficient Fine-Tuning (PEFT) using LoRA.
• Applied Reinforcement Learning from Human Feedback (RLHF) concepts in PyTorch, including reward modeling, Proximal Policy Optimization (PPO), and challenges such as reward hacking.
• Experimented with model optimization techniques quantization (bfloat16), prompt design, reasoning strategies (Chain-of-Thought, ReAct), and external API integration.

Backend Development & Security
• Built a Spring Boot web application with JPA/Hibernate, designing a MySQL relational schema and integrating persistence using Spring Data JPA.
• Implemented Spring Security with JWT to secure REST APIs, enabling role-based access control and stateless authentication.
• Applied AES encryption and tested multiple modes of operation (CBC, GCM, etc.) to protect sensitive data.
• Developed a secure full-stack Java web app using JDBC, Servlets, MySQL, and jQuery (AJAX), improving authentication and session management with SSL encryption.
• Explored and applied ClickHouse for large-scale data analytics, including query optimization, schema design and analysis for efficient data processing.

Systems, Databases & Networking
• Designed and optimized a Microsoft SQL Server database, integrating stored procedures for efficient data handling.
• Implemented multi-threaded client-server architecture in Java sockets, supporting concurrent users with secure data transmission.
• Enhanced debugging and performance by implementing Log4j2 logging and analyzing network traffic with Wireshark.
Tools & Documentation
• Worked on the company’s technical documentation using Markdown and AsciiDoc with custom themes.
• Created PlantUML diagrams to clarify system architecture and application logic.