2023-09-22, 11:30–12:00 (Europe/Bucharest), Room 1
Question-answering and summarization; Contextual information extraction; Integrating Large Language Models
In this presentation we will discuss a technique for solving the need of interactive question-answering and summarization of documents using the latest developments in the Machine Learning field (Large Language Models). We will explore open source technologies like LangChain, LLM (Falcon / Mosaic MPT / LLaMA from HuggingFace platform), techniques like Prompt Engineering in order to achieve the integration between Libre Office documents and state of the art NLP models.
Irina graduated Computer Science in Polytechnics University of Bucharest in 2007 and started her career in the last year of University as a .NET developer. Then her path moved to Java developer, team lead, delivery manager and now to head of development in 1&1 Internet Development. As head of development in 1&1 Internet Development she leads a team of 9 developers and they are building 3 of the products from the portfolio of the portals (mail.com, gmx.com, gmx.de, web.de) that our company is managing.
Irina is involved in the Online Office product development where her team is integrating Libre Office and Collabora Online into the web.de / gmx.de / gmx.com / mail.com portals.
Andrei Mihai is a passionate software developer and tech enthusiast at 1&1 Internet Development. He has spent over 16 years developing Java applications, microservices, DevOps pipelines and in the last years, Machine Learning tools.
Andrei is also very interested in Quantum Computing and finding real usecases for applying Quantum Technologies.