As alternative datasets mature and become embedded within investment workflows, the distinction between “market” and “alternative” data is fading. Investment teams are increasingly sourcing and building integrated data ecosystems that combine traditional sources, proprietary research and alternative signals. This panel explores the challenges of this convergence; from integration and governance to scalability and unlocking value from increasingly diverse datasets.

In this showcase session hear from some new entrants to the market data space, as they share the datasets and product launches set to challenge established vendors on quality, flexibility (and price!).

Compute is becoming a financial asset, but the market pricing it still trades largely in the dark. This session covers the infrastructure making it tradable: GPU forward curves, residual value models that replace straight-line depreciation with discounted rental income, independent performance benchmarking, and the indices soon settling CME's compute futures. Silicon Data CEO Carmen Li walks through what a reference price layer for AI compute looks like and how it is built.

How can investors identify emerging retail-driven short-squeeze risk before it becomes obvious? Travis Whitmore discusses how securities lending, digital media, and social sentiment data can be combined into a useful signal for detecting crowded positions and potential market dislocations.

Market data has always been the foundation of quant investing, but AI is exposing where that foundation needs to change. This session examines how quant teams are rethinking their data stack to support AI at scale: real-time vs. historical architecture, tick-level granularity, and vendor consolidation. Speakers will discuss what's actually changing in how market data is sourced and delivered for AI-driven investing.

As AI transforms investment research, access to information and to the tools that process it is being rapidly democratised. But not all data, and not all research, is created equal. As the analysis layer commoditises, a harder question comes into focus: where does edge actually come from when everyone has capable AI? This panel brings together buy-side and vendor perspectives to examine what still generates alpha in 2026. We want to explore which datasets, sources and vendor relationships remain genuinely hard to replicate, how buyers distinguish real differentiation from repackaged public information, and the extent to which human analyst judgement, the forward-looking insight and institutional knowledge that models can summarise but not originate, still carries measurable value. Drawing on Neudata's cross-market view of hundreds of providers, the session aims to separate what is genuinely proprietary from what has already been arbitraged away, and to map where the next wave of differentiation is likely to emerge.