Eya Dridi

◦ Developed an end-to-end AI pipeline for automated evaluation of requests for proposals (RFPs) and generation of structured proposals using OpenAI’s LLM models. ◦ Created an advanced data ingestion module for processing complex documents, using OCR for scanned files and a multimodal LLM (Llama) for content extraction from PowerPoint presentations. ◦ Implemented a Retrieval-Augmented Generation (RAG) architecture with LangChain, using PostgreSQL and pgvector for vector storage and similarity search. ◦ Integrated a feedback loop based on user corrections to continuously improve the accuracy of future generations, and exposed the solution through a REST API developed with Flask.

◦ Developed an end-to-end generative chatbot based on a RAG architecture to automate customer support, integrating Google’s Gemini LLM. ◦ Designed a complete data pipeline, from web scraping with BeautifulSoup to generating embeddings and indexing them in a FAISS vector database for efficient semantic search. ◦ Developed a REST API with FastAPI for the backend and an interactive user interface with Streamlit, integrated as a widget. ◦ Containerized the multi-service application with Docker to ensure a modular and scalable environment.

Contributed to the design and development of a dynamic e-learning platform using React.js. The platform offers a seamless and interactive user experience, leveraging React Router for smooth navigation and Ant Design for a clean, responsive interface.