Abstract: 

With the development of Transformer-based language models, NLP has had a leap in research and techniques in the last few years. These language models enable different tasks such as Question Answering, summarisation, translation, retrieval and so on. These, combined with vector optimized databases and the development of Open Source frameworks such as Haystack have made it possible for us to create NLP powered applications to a quality that was previously not possible. This talk will cover an intro to NLP and Question Answering, followed by an example on how to build a Question Answering pipeline with Haystack.

Session outline - What you'll learn:
-What is the current state of NLP for Question Answering
-Why is Semantic Search so relevant here?
-Learn the structure of the Haystack framework to build a search system
-See how the framework Haystack, together with a Vector Database like Weaviate, can build a Question Answering pipeline

Bio: 

Tuana is a Developer Advocate at deepset. She works on improving the developer experience and adoption of deepset's Open Source NLP framework: Haystack. Originally from Istanbul, she moved to the UK in 2014 where she obtained a Master's degree in Computer Science from the University of Bristol (in 2018). She initially started her career as a Software Engineer but then decided to become more involved with open source communities and educating people. This led her to developer relations. She worked as a developer advocate at Cumul.io before moving to deepset in 2022.