ChatVitae

About the Project

I developed ChatVitae as a personalized digital representative of my professional identity—an interactive AI that allows visitors to converse directly with my CV. Using a Retrieval Augmented Generation (RAG) architecture, the system intelligently retrieves relevant segments of my resume when answering user queries, ensuring responses are both accurate and contextually appropriate. The application leverages LangChain's document processing capabilities to chunk and index my CV, while OpenAI's embeddings create semantic representations that power the retrieval mechanism. What makes ChatVitae special is its conversational memory—it maintains context throughout discussions, remembers user names, and responds with a consistent persona that represents me professionally. Initially built on GPT-3.5, I've since upgraded to GPT-4o-mini for improved response quality while maintaining performance. The entire system is containerized and deployed on Google Cloud Run for reliability and scalability.