When
Thursday, November 7, 2024 - 4:00 p.m.
Felipe Viana
Head of Applied Research
Articul8
"Building Expert-level Generative AI Applications in Engineering – Challenges and Opportunities"
AME Lecture Hall, Room S212
Zoom link
Abstract: Enterprises leveraging general purpose Generative AI (GenAI) models and platforms struggle to progress beyond initial experimentation and proof-of-concept projects focused on information search and retrieval. This is mainly because enterprise-grade GenAI applications need to consider a myriad of other factors such as data security, domain-specificity, personalization, scalability, auditability, price-performance optimization and more, to accelerate outcomes and return on investment. If companies fail to overcome these challenges and halt their GenAI adoption, they risk slowing down their innovation capacity, falling behind competitors who successfully harness the transformative potential of GenAI, and ultimately, compromising their long-term competitiveness. This is the reason why implementing successful enterprise-grade GenAI applications is a non-trivial endeavor that cannot be solved using many of the existing methods and technologies. This presentation explores the challenges and opportunities associated with building enterprise-specific GenAI applications. We will introduce a novel approach that enables companies to successfully deploy and scale enterprise-grade GenAI applications. Our approach leverages a modern, secure, vertically-optimized enterprise-grade GenAI platform, specifically designed to bridge the gap between proof-of-concept and production-ready deployments. We will illustrate the potential of this approach through showcases in manufacturing and in aerospace, highlighting the opportunities and benefits of expert-level GenAI applications in engineering.
Bio: Dr. Felipe Viana is currently the head of applied research at Articul8, where he oversees a team of researchers and engineers developing Generative AI solutions that meet the needs of Articul8's clients. In addition, Dr. Viana is responsible for establishing the research roadmap and new technology introduction programs; as well as building research partnerships with academic institutions and external partners. Prior to this role, he held a series of senior positions, including director of artificial intelligence at Intel Corporation, where he defined and prioritized product roadmap and strategy, built early pilot programs, and supported the deployment of effective and targeted go to market activities. Dr. Viana also served as an assistant professor of applied machine learning at the University of Central Florida, establishing a research program on scalable machine learning algorithms and teaching courses on multidisciplinary optimization, quantification of uncertainty, and applied probabilistic methods. Additionally, he worked as a senior scientist at GE Renewable Energy, developing algorithms to improve asset predictability and performance, and creating tools for remaining useful life prediction. Earlier in his career, Dr. Viana conducted research on multidisciplinary optimization, design under uncertainty, and engineering reliability at GE Global Research, applying their work to various GE systems.