souissi ons

Developed a pre-consultation tool that leverages conversational AI and RAG for symptom assessment. Data Engineering: Built Python web scrapers to extract data from authoritative medical sources. Designed pipelines for cleaning, normalizing, and transforming data (XML/HTML/CSV/JSON). Created condition-symptom knowledge bases and conversational datasets. AI Model Development: Trained a custom GPT-2-inspired model in PyTorch. Fine-tuned LLMs on domain-specific dialogues using PEFT (LoRA) and 4-bit quantization for resource-efficient training. Used Hugging Face (Transformers, PEFT, TRL, Datasets) for model workflows. Backend (FastAPI): Designed an MVC-style REST API with endpoints for ingestion, processing, and semantic search. Used Factory pattern to flexibly support embedding models and vector DB. RAG Pipeline: Built an end-to-end RAG system with LangChain and LangGraph. Indexed medical texts via chunking, embeddings, implemented semantic search to inject relevant context into LLM prompts. Frontend (Streamlit): Developed an interactive UI supporting multi-turn dialogues, symptom elicitation, user consent, and pre-consultation report generation.

In-Depth Lead Analysis: Conducted comprehensive research to evaluate the viability and market potential of potential leads. High-Quality Lead Sourcing: Identified and compiled targeted leads using online platforms and databases, ensuring relevance and accuracy. Strategic Insights: Provided actionable recommendations to optimize lead acquisition strategies.

Data Collection: Utilized Playwright, Scrapy and BeautifulSoup for web scraping from Bayut website. Data Processing: Conducted data cleaning and ETL with Pandas and NumPy. Data Storage: Designed schema and managed structured data in MongoDB. Visualization: Developed an interactive real estate dashboard in Mokkup.ai and Streamlit for UAE property metrics.

Comprehensive Testing & Quality Assurance: Conducted extensive functional testing across all product features within production environments, ensuring high performance and reliability. UI/UX Enhancements: Proposed and implemented user interfaces and frontend improvements. Automation & Development Optimization: Automated daily test reporting by extracting data from Trello and generating detailed reports, while designing and executing test scenarios using Gherkin syntax in Squash to streamline behavior-driven development.

Data Management and Analysis: Conducted market research and company profiling to support partnership strategies and project planning, utilizing data to inform decision-making and improve outreach effectiveness. Project Management: Led cross-functional teams to implement initiatives that promoted international youth exchanges, aligning with AIESEC’s mission and impact goals. Event Organization: Coordinated large-scale events and conferences focused on leadership development and cultural exchange, enhancing the AIESEC experience for local and international members. Team Development: Mentored and supported team members to achieve personal growth and performance targets, contributing to a positive and high-performing team culture.

End of studies professional project (IOT project) Realising a smart medical bracelet that allows the measurement of physiological parameters with an alert in case of danger. Developing a platform to ensure communication between doctors and patients