Yassine Tahiri Alaoui

– Designed and deployed production-ready autonomous multi-agent systems using SecureGPT APIs for real-time secret detection and vulnerability classification across code repositories and logs. – Implemented core agentic capabilities including adaptive reasoning (Chain-of-Thought), memory management, and tool orchestration to automate security risk prioritization. – Fine-tuned Large Language Models (LLMs) and hybrid deep learning architectures to achieve high-accuracy classification in security-critical environments, significantly reducing false positives. – Developed scalable, low-latency API endpoints using FastAPI and ensured seamless deployment via Docker for large-scale enterprise integration. – Built custom NLP pipelines for precise extraction and filtering of potential credentials, aligning agent behaviors with proactive defense against OWASP-related exposures. – Created interactive dashboards using React and Next.js to monitor agent behaviors, classification performance, and detection outcomes.

– Analysis of Attijariwafa Bank's Call Centers Recordings Using Speech-to-Text Technologies – Preprocessed audio data using SpeechBrain to enhance recording quality. – Fine-tuned the OpenAI Whisper Speech-To-Text model according to specific requirements. – Implemented topic modeling of transcriptions using zero-shot classification. – Developed sentiment analysis using scikit-learn. – Created a user interface with Tkinter for real-time analysis of customer interactions. – Developed interactive and predictive dashboards using Power BI.