Pneuma-Seeker is an agentic system that helps data analysts refine vague questions by converting them into explicit, inspectable relational specifications. Leveraging LLMs as transparent, interactive collaborators, it enables iterative query refinement and data discovery. Two real-world procurement case studies demonstrate its effectiveness for practical data exploration workflows.
Research
Demonstration of Pneuma-Seeker: Agentic System for Reifying and Fulfilling Information Needs on Tabular Data
Pneuma-Seeker demonstrates how agentic systems can iteratively refine vague data analysis questions into executable relational specifications via interactive LLM collaboration, enabling practical procurement data exploration workflows.
Friday, April 17, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.AIBY sys://pipeline
Tags
research
/// RELATED
Strategy4d ago
More than 90,000 tech workers have been laid off this year. But here’s why companies like Microsoft are offering voluntary buyouts instead
Meta and Microsoft are shifting from mass layoffs to voluntary buyouts as AI spending pressures force selective workforce restructuring—Meta's 10% cut leaves 6,000 unfilled roles while Microsoft offers buyouts to 8,500 U.S. workers.
InfrastructureApr 28
Nvidia’s Jensen Huang thinks $1 trillion won’t be enough to meet AI demand—and he’s paying engineers in AI tokens worth half their salary to prove it
Jensen Huang signals Nvidia's $1 trillion forecast is still short of peak AI infrastructure demand, so the company is paying engineers half their salary in compute tokens to lock in talent during the infrastructure arms race.