SIGdial, Hello. I write, today, about a new BudgetQA challenge for dialogue systems R&D. This new challenge hopes to enable the advancement of AI-enhanced budget analytics. Numerous societal benefits would result from AI-enhanced budget analytics. Budget committee members and their staffs would be able to better engage with budget proposals and versioning budgets. Citizens (e.g., journalists) would be able to better engage with governmental budget data. To technical topics, we can envision multi-hop processing for answer production. We can envision, beyond simple exchanges, multi-step dialogues. We can envision, beyond text dialogues, multimodal dialogues where answers might include information visualizations such as pie charts. We can envision, beyond dialogues about single budgets, multi-budget dialogues where questions and answers could be about multiple budgets, e.g., different years of an annual budget, simultaneously. A combination of multimedia-enhanced dialogue system responses with multi-turn dialogue suggests an opportunity to provide end-users with features for asking follow-up questions involving selections of machine-generated charts, diagrams, figures, graphs, or infographics. Technical topics also include providing a capability for end-users to be able to save and load dialogue state data into and from URL-addressable slots. With such a capability, end-users could, using a graphical user interface or via dialogue, request to receive a URL for themselves and others to use to enter saved dialogues at points of interest. Points of interest in multimodal dialogues could, for instance, have complex data visualizations displayed on screen. Thank you. I hope that this new challenge domain, BudgetQA, is of some interest to you! Best regards, Adam Sobieski <http://www.phoster.com>