This presentation will first introduce the concept of Large Language Models (LLM) and highlight the strengths of Open Weights LLMs. Following the introduction, the concept of how to effectively use these models will be showcased through a use-case specifically for patients diagnosed with ADHD. Ethical and privacy concerns will be also addressed for this approach and eventually the session will be concluded with a demo that will showcase how open-weight LLM bots can be created using open source software like Ollama, Llama.cpp, LlamaIndex and Streamlit, and deployed with Kubernetes/Openshift.
Kyra Goud is a Specialist Solution architect based in Switzerland with a strong background in NLP and Computer Vision. With a data scientist background and approach, she is actively engaged in bringing the MLOps pipeline closer to customers.
Rodeina Mohamed is a Technical Graduate at Red Hat based in Munich with a strong background in Data Science and Software Engineering. She is interested in bringing AI applications to the cloud and pioneers for open source AI applications.
