Generative AI Use Cases

Sponsored by SAS (www.sas.com).

--- Talk #1 ---

Tamara Fischer, Principal Data Scientist, SAS

Synthetic Data Generation: The Art of Balancing Usability and Privacy
Synthetic data generation, a subset of Generative AI (GenAI), is revolutionizing data science by creating artificial data that mimics real-world data. This approach is essential for several reasons: it helps make data accessible, protect data privacy, mitigates biases, and augments data required for training machine learning models. Typical use cases include life sciences, finance, public sector & manufacturing, where synthetic data can add value. Synthetic data can be generated using different methods such as simulation, statistics or GANs (Generative Adversarial Networks). However, important aspects to consider include ensuring data fidelity, maintaining privacy, and addressing ethical concerns. Balancing these factors is crucial for leveraging synthetic data effectively while safeguarding privacy. In my presentation, I’ll provide an overview on how to measure privacy for synthetic data including a practical example.

Bio: Tamara Fischer, a graduate statistician, has been working for many years in the role of Principal Solutions Architect, Analytics, at SAS in the DACH region. In this role she has implemented solutions along the entire analytical life cycle from model development to model deployment. Currently she works in an international team consisting of data scientists who take care for the EMEA region. Her work comprises all analytical aspects with a focus on operationalizing analytical use cases and responsible AI.

--- Talk #2 ---

Clément Lefebvre, Senior Data Scientist, SDSC

Generative AI at SDSC: Advancing Knowledge, Extraction and Creativity

Bio: Clément Lefebvre is a Senior Data Scientist at the Swiss Data Science Center (SDSC). He holds a Master of Science in Computational Science and Engineering (2018) from EPFL and joined the SDSC in 2019, focusing on industry and applied research projects. Clément has contributed to multiple projects leveraging Large Language Models (LLMs) for innovation, notably developing AI-driven solutions for the public sector, NGOs, and various industries.

--- Talk #3 ---

Marcelo Yannuzzi, Principal Engineer, Cisco

Controlled inference glueing language models, enterprise data, identity, applications, and location

Bio: Marcelo is with Cisco’s innovation arm, where he is leading the design and development of innovative technologies bringing together GenAI, enterprise data, identity, applications, and agents with focus on secure and compliant utilization of GenAI in novel AI fabrics. Marcelo has 70+ patent applications, 100+ scientific publications, several technical publications, and editorials in the scientific community, including a Cisco Press Book.