Six sourced rows, one per player: Meta’s acquisition of Sundial (announced 23 October 2026), Sundial’s own platform, and one current official offering each from Google, Microsoft, Amazon, and OpenAI — each with its documented strength and its documented gap, from six distinct sources.
Rows are the capabilities of a production-grade analytics trust layer; columns are vendors. Hover or tap a cell for the sourced detail behind the rating. Ratings are an editorial reading of each vendor’s own documentation — sources linked below.
Interpretive matrix: ratings summarize each cited source; “not stated” means the source does not claim it, not proof of absence. The buy-rather-than-build conclusion is a strategic inference, not an announced deal by any competitor.
The acquisition profile competitors would need to replicate or buy. Sundial stops serving external customers when the Meta deal closes.
Captures business context, domain knowledge, semantics, and the company’s analytical model — the layer that makes answers mean what the business means.
Analytical workflows written by domain experts, so the agent analyzes the way the organization’s best analysts do.
A confidence signal on every answer, distinguishing governed definitions from ad-hoc SQL.
Every step transparent and auditable — the control plane that lets a regulated enterprise trust an agent.
Evaluations that catch quality regressions — and all of it deployed over existing warehouses, BI tools, and LLMs, vendor-neutral.
| Entity | Offering / event | Documented strength | Documented gap or advantage | Source |
|---|
Method: six-row single comparison set assembled from vendor documentation and press coverage (one source per row, six hosts, retrieved around the October 2026 Meta–Sundial announcement). Strength and gap texts are condensed quotes from each source; matrix ratings are an editorial reading of those texts on a three-step scale (documented / partial / not stated). Long source passages trimmed for space; full context at the linked pages.