ACHRAF HABIB

Context: As part of an initiative to streamline ICU patient monitoring, I developed a user-friendly web application and an MLOps pipeline, enabling healthcare employees to easily track and predict patient complications in real-time. Achievements: ● Training Deep Learning models for predicting medical complications, versioning and logging models. ● Design and implementation of an MLOps pipeline to manage the models lifecycle. ● Development of a web app for real-time patient data predictions. ● Automating testing and deployment through a CI/CD pipeline. ● Containerization and deployment of the application ● Monitoring of the application by tracking model performance and stability.

Context: As part of a digital transformation initiative aimed at improving access to data and documentation, I developed a conversational chatbot to facilitate the use of Vermeg's internal framework, Palmera, which is based on the Spring framework. The chatbot was designed to simplify learning and usage for employees, interns, and apprentices. Achievements: ● Design and implementation of a conversational chatbot ● Langchain RAG Integration: Integrated Langchain's RAG framework to combine retrieval and generation techniques for enhanced performance. ● Training and Fine-Tuning the chatbot for diverse conversational contexts and accurate responses. ● Retriever Implementation: Added a retriever component to fetch relevant documents, improving context-aware replies. ● Management of vector embeddings and optimization of retrieval performance through efficient similarity searches. ● Development of a conversational chain for coherent dialogue flow. ● Web app Development: Created a Streamlit web app for real-time chatbot interactions, ensuring a user-friendly interface.