Hi there, I'm Máté Boros 👋

• Education & Knowledge Transfer: Served as a professional mentor and instructor within the strategic partnership between the company and Széchenyi István University. Responsible for teaching AI development, practical modeling, and modern architectures (e.g., RAG) to students, facilitating the integration of corporate technologies into the academic curriculum.
• Intelligent Tender Analysis System (RAG): Developed an LLM-based decision support system utilizing RAG architecture to automate the processing of complex tender documentation. The system enables natural language querying and provides automated scoring based on fleet capacity and route optimization parameters, significantly accelerating the workflow for freight planners.
• Computer Vision-Based Damage Assessment System: Implemented a site entry support solution by fine-tuning RF-DETR and YOLO models. The system is specifically optimized for the real-time detection of aesthetic defects, scratches, and structural damage on trucks and heavy-duty vehicles. https://www.portfolio.hu/gazdasag/20260527/napi-25-ora-munkat-sporol-meg-a-ai-a-waberers-nel-838266
• Document Digitalization Ecosystem: Developed custom OCR and document analysis models for automated data extraction from vehicle registration certificates and internal documents. Built a searchable knowledge base from image-based sources, minimizing administrative overhead and manual data entry errors.
• Internal Knowledge Management Chatbot: Developed a RAG-based AI assistant powered by internal corporate data, providing employees with instant access to company-specific information, guidelines, and regulations.

Azure AI Foundry and React based web development. Developing a custom time-tracking interface from ground-up with automatic Teams Meeting calendar integration to improve data accuracy.

CRM development using React for automated DOCX contract filling. Proficient in Firebase database management. Project modernization, including Framer Motion-driven UI/UX and performance enhancements. Building Go/Python APIs for Google Cloud Run.