[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"article-body-what-is-a-decision-layer":3},"\nA decision layer is the tier of the enterprise stack that sits\nabove the systems of record and beneath the people who decide. It\nturns the data an organisation already holds into predictions with\ntested odds, early warnings routed to named owners, plans tracked\nagainst outcomes, and answers computed inside the asker's\npermissions. A system of record documents what happened. A\ndecision layer changes what happens next, and keeps the score that\nproves it.\n\nThe term exists because a gap exists. Twenty years of enterprise\nsoftware produced superb machinery for\n[recording, tracking and reporting](/insights/the-maturity-ladder-ends-at-prevention),\nand\n[records are not decisions](/insights/records-are-not-decisions).\nThe judgement that actually runs the business still lives in\nspreadsheets, inboxes and meetings, one floor above the systems\nthat were supposed to inform it. The decision layer is the name\nfor the tier that closes that gap.\n\n## The four requirements\n\nA system qualifies as a decision layer when it meets four\nrequirements, and each one is checkable.\n\n- **Predictions with published odds.** Every claim about the\n  future carries a probability that has been tested against what\n  actually happened, so that\n  [of everything called 70%, about 70% lands](/insights/of-everything-called-70-percent).\n- **Deterministic numbers.** The\n  [same inputs produce the same figure, every time](/insights/same-inputs-same-number),\n  so two people can reconcile what they saw and an auditor can\n  regenerate it.\n- **A named owner and a threshold.** Every material signal reaches\n  one accountable person with a deadline, so\n  [millions of events become a handful of decisions](/insights/millions-of-events-in-a-handful-of-decisions-out)\n  rather than a feed.\n- **A record of outcomes.** Every call is scored against what\n  happened next, so the organisation learns which of its\n  predictions, and which of its people's overrides, deserve trust.\n\nTwo properties hold the four together. Everything runs on\n[one event backbone](/insights/one-event-backbone), so a\nprediction, the intervention it triggered and the outcome it\nproduced stay connected end to end. And every number is computed\ninside the viewer's entitlements, so each person sees\n[their own honest total](/insights/one-number-per-viewer) and\nnothing beyond it.\n\n## What a decision layer is not\n\nThe boundaries define the term as much as the requirements do.\n\nBusiness intelligence describes. It answers what happened, in\ncharts, and leaves what to do about it entirely to the reader. A\ncopilot on a system of record\n[summarises the past faster](/insights/your-copilot-is-answering-the-wrong-question),\nwhich is useful and is not deciding. Workflow and process tools\nroute tasks between people; they move the work without predicting\nanything or scoring what was decided. And a rules engine does\ndecide, but without odds, calibration or learning, so it cannot\ntell you how often it has been right.\n\nNone of these are failures. Each does its job. But a dashboard\nwith predictions bolted on, a chat window over records, or a\ngovernance platform standing beside the stack does not become a\ndecision layer by adopting the word. If the odds are untested, the\nnumbers change on refresh, nobody owns the exception, or no\noutcome is ever scored, the layer is missing exactly where it\nmatters.\n\nAnalysts file this territory under Decision Intelligence, and the\nshelf is useful for finding vendors. The label undersells the job.\nInsight is the entry fee; a decision layer's product is owned,\nscored outcomes.\n\n*A system of record is complete when it describes the past. A\ndecision layer is complete when the next outcome improves.*\n\nProphesee is built as exactly this layer, with the four\nrequirements as\n[the architecture rather than the aspiration](/platform).\nWhere records become decisions.\n[See it on your own data](/contact).\n",1786984938090]