Latavis is built as a single prediction gateway with workspaces, an AI layer, and spreadsheet formulas on top. The part that makes it different: a verification envelope — level, confidence cap, as-of, source — attached to every value, everywhere it travels.
Cross-asset view: the flagship recession probability, macro conditioning, validated-signal tiles, morning-brief teaser, and alerts — all re-colored by the global scenario lens (base / bull / recession / tail).
A uniform workspace per vertical: national picture, entity explorer, per-entity predictions with the full honesty envelope, and data health. Private Credit ships the richest workspace — a real EDGAR BDC panel with LGD and ASC-820 Level-3 marks.
Roll positions up across verticals, run scenario P&L, propagate shocks between markets, and use the portfolio optimizer — under the same scenario lens as the rest of the platform.
A scenario runner for stress paths across your exposures, with results that carry the same verification badges as everything else.
Chat, brief, and memo answers cite their sources inline, with tool traces you can inspect. Citations aren't a feature; they're the admission requirement.
A conversational analyst over the whole platform. Every answer shows inline citations and an inspectable tool trace — you can see exactly which endpoints and data produced each claim.
A generated cross-market brief, cited end-to-end, ready when you sit down.
Pick an entity, a vertical, and a focus — get a structured, cited memo you can hand to a committee.
A browsable catalog of =PRISM() functions with a live evaluator that returns value, verification level, confidence, and provenance — plus an Excel add-in, because your model lives in a spreadsheet.