i am a 1st-year economic informatics student in romania, deeply passionate about low-level system architecture, computer vision, and offensive/defensive security. i am transitioning from full-stack web development to backend engineering, focusing on how operating systems interact with memory and networks at the lowest levels.

i don't just write scripts; i reverse-engineer problems, analyze system behaviors, and build custom tools to automate, intercept, or optimize processes.

• windows internals & reverse engineering: deep understanding of ring 0 vs ring 3 privileges. familiar with kernel-level driver interactions, memory management, direct memory access (dma) concepts, and api hooking techniques.
• network security & traffic analysis: building tools for deep packet inspection (dpi). i understand how to intercept, parse, and block network traffic (tls/sni, dns) before it leaves the windows network stack using wfp (windows filtering platform).
• computer vision & autonomous systems: designing state-machine-driven agents that process screen data in real-time. experienced with hsv color space masking, morphological filtering, and simulating low-level hardware inputs to bypass software restrictions.
• process automation: extracting and manipulating unstructured data using python, computer vision, and ai endpoints to build enterprise-grade data pipelines.

here is the code that backs up my research:

• aegis: kernel-level network firewall | python, windivert, networking a custom threat-interception engine. it hooks into the network stack to manually parse raw tls client hello packets and dns queries, blocking c2 infrastructure and telemetry without decrypting the payload.
• cv-autonomous-agent | opencv, numpy, win32api an intelligent computer vision agent capable of real-time object detection and decision making. it uses screen-grabbing (mss), visual blockage detection, and kernel-level mouse events (ctypes) for error recovery and dynamic navigation.
• Facturino | python, sql, ai apis an automated business solution that connects local scripts with ai to parse complex, unstructured invoices (pdf/jpg) and structure the data into relational databases and excel reports. work in progress
• reclame pe paine | javascript es6, css3, html5 a production-ready, mobile-first commercial platform built for a marketing agency. implements intersection observer for performance and strict client-side validation for secure lead generation.

• languages: python, c/c++ (reading/research), javascript (es6), sql, html/css.
• low-level & networking: win32api, ctypes, windivert, socket programming.
• vision & data: opencv, numpy, mss, pandas.
• tools: git, vscode, visual studio.

i am actively looking for an internship or junior role in backend development, cybersecurity, or automation. i am eager to bring my problem-solving mindset to a team that builds serious software.

• email: [d19691551@gmail.com]