Ad-hoc Data Exploration using Conversational AI
Abstract: Conversational AI is getting more and more widely used for customer support and employee support use-cases. In this session, I'm going to talk about how it can be extended for data analysis and data science use-cases, i.e., how users can interact with a bot to ask ad-hoc analytical questions on data in relational databases. This allows people to explore complex datasets using a combination of text and voice questions, in natural language, and then get back results in a combination of natural language and visualizations. For example, they can ask questions like "How many cases of Covid were there in the last 2 months by state and gender" or "Why did the number of deaths from Covid increase in May 2022", and get back results in a combination of text and charts.
Presenter: Anand Ranganathan is a co-founder and the Chief AI Officer at Unscrambl. He is leading Unscrambl's product development in natural language processing, automated insights and data story-telling. Before joining Unscrambl, he was a Master Inventor and Research Scientist at IBM. He received his PhD in Computer Science from UIUC, and his BTech from the IIT-Madras. He also has over 70 academic journal and conference publications and 30 patent filings in his name.
Human-AI collaborative knowledge discovery in tradespace exploration
Abstract: Tradespace exploration is a method used in the early design of complex systems to explore a wide range of design alternatives across conflicting figures of merit in a rigorous and systematic way. AI tools have been used in tradespace exploration for decades, primarily to improve efficiency in the evaluation and search of the design space. However, what engineers need from AI is more than just better design optimization/search capabilities to find the best possible designs according to the models. Engineers need to learn what is driving those results, why they are getting those results, and under what circumstances they remain valid, so they can justify their design decisions to stakeholders. In summary, the goal is insight, a rather subjective and ill-defined construct that has been studied in the visual and data analytics community for decades. In this talk, I advocate for the use of AI agents to help engineers explore the large and complex datasets resulting from tradespace exploration in the context of their own knowledge and expertise of the problem. I will summarize a few studies we have done in my lab to better understand what humans learn in design space exploration, how AI assistants can maximize that learning, and the subtle relationship between how much engineers learn about a problem and other more traditional outcomes of tradespace exploration such as the quantity, quality and diversity of designs explored.
Presenter: Daniel "Dani" Selva is an Assistant Professor of Aerospace Engineering at Texas A&M University, where he directs the Systems Engineering, Architecture, and Knowledge (SEAK) Lab. His research interests focus on the application of knowledge engineering, global optimization and machine learning techniques to systems engineering and design, with a strong focus on space systems. Dani has a dual background in electrical engineering and aerospace engineering, with degrees from MIT, Universitat Politecnica de Catalunya in Barcelona, Spain, and Supaero in Toulouse, France. Before his PhD, Dani worked for four years in Kourou (French Guiana) as an avionics specialist within the Ariane 5 Launch team, where he successfully launched 21 Ariane 5 rockets to space. He is a Senior Member of AIAA and IEEE, and the Secretary of the AIAA Intelligent Systems Technical Committee.
More details: www.incose.org/ai