Value Engine, ValueAtlas by Auricita

Value Engine

Turns the report's recommendations into a tailored activity framework that surfaces the value.

ValueAtlas is the full lifecycle tool behind Value Mining, driving everything from intake to the value it ultimately realizes. That lifecycle runs in two halves. The first is the analysis, covered on the Watch a demo page: intake, the interview, the analysis run, and the report. The second starts the moment that report is delivered, and it's what Auri calls the value engine: a rolled-up dashboard that tracks recoverable value as it's actually earned, metrics that Auri watches, and a phased plan that tracks whether the recommendations are actually being acted on.

What you track

The running number that keeps moving after the report

The Recoverable Value Dashboard

This is what keeps moving after the report is delivered: one running total across every analysis a person can see, updated the moment your own numbers move, not on the next scheduled review.

One rolled-up total, scoped to who's looking. A single dashboard tracks recoverable value across every analysis you can see, not just the one you most recently ran; an owner sees the whole organization's total, a scoped team member sees only their own processes, from the same screen.
Dollars tied to named value levers, not spread arbitrarily. Each dollar figure is tied to one specific value pool ("lever") from the report, and the metrics that move are shown as the drivers proving that lever, so the money you see and the metrics you see are guaranteed to add up and never contradict each other.
No credit until something is proven, but real credit once it is. The recovered-value bar carries no fill at all until a real reading confirms movement; there is no minimum sliver implying progress that hasn't happened. Once a reading does confirm movement, partial progress counts for real, proportional dollars rather than waiting for a target to be hit outright.
A "Confirmed by" line on every priced value row. Each dollar figure names the one metric whose actual movement in your real results proves that value is being realized, or says "No metric confirms this row yet" when nothing has been measured, so every dollar claim on screen is backed by a specific, checkable number rather than an assertion.
Unpriced findings listed separately, and fixable right there. A recommendation still waiting on a measurement is listed by name at the bottom, clearly separated from priced rows above so you never mistake "found but not yet priced" for "already recovered," and if it's missing a cost or volume figure, you can supply that number right on the dashboard and it reprices immediately.
Admin control over what counts toward the total. An organization can exclude a repeat analysis of the same process from the total, so re-running an analysis never double-counts value already claimed.
Signed tradeoffs shown honestly, not flagged as failures. When a recommendation deliberately accepts a small, intentional loss in one metric to produce a bigger gain elsewhere, the dashboard shows both sides explicitly (the gain, the accepted cost, and the net), so a metric that's expected to move slightly the wrong way on purpose is never mistakenly shown as at risk.
A scheduled digest to whoever owns the outcome. A recurring summary of recoverable value can go out on the cadence you choose (weekly, monthly, or off), sent to whichever person you designate, with links straight back into the relevant analyses; if no one is specifically assigned, it falls back to your organization's admins automatically, so a digest is never silently unsent.
The same numbers everywhere. The dashboard, the Metrics view, and the Progress view all read the same underlying figures, so a dollar amount can never disagree depending on which screen it's read from.

What Auri tracks for you

Ongoing monitoring, without anyone having to go looking

Metrics and Targets

Metrics is whether the numbers are moving. Layered on top of it, Auri watches those same numbers on her own and raises what changed, what to worry about, and what waiting is costing, in plain business language rather than a raw statistical alert.

A real trajectory, not a single percentage. Every tracked metric shows baseline, target, and current value plotted together, with a percent-to-goal figure and a status color, so progress reads as a line, not just a snapshot.
Every baseline and target names its own source. A figure pulled from a customer's documents says so; a figure still missing prompts the customer to supply it rather than sitting blank.
Update one number, or refresh them all at once. A single reading can be typed in with its date, or a whole set of metrics can be refreshed in one pass through a bulk upload, using a template pre-labeled with that organization's own metric names.
Corrections stay on the record. Entering a new value for a date already on file replaces it; a reading can be removed individually if the date itself was wrong, but nothing is ever silently discarded.
A single health score for the whole process. A 0 to 100, dollar-weighted score for progress toward target, shown with the trend since the last check and exactly how much of the tracked data it's based on.
A stated cost of waiting. A dollar range for what leaving the process unfixed costs per year, so the case for acting is quantified before anything has moved yet.
Honest status coloring: green is earned, not assumed. A metric is only shown "holding" once it has actually reached its target; one that's improving but not there yet gets a distinct "on track, still short" color, one drifting the wrong way is flagged clearly, and one with no reading yet is marked as awaiting, so you're never shown a false "all clear" just because a number exists.
A ranked, plain-language queue of what needs attention. Auri surfaces metrics drifting past target, metrics that just hit target, overdue re-measurement checkpoints, and findings still waiting on a price, capped and ranked by how much value is at stake so the most important ones show first.
Tunable, not all-or-nothing. Any metric's sensitivity can be turned down so only bigger changes get flagged, or muted entirely, and muting clears it everywhere at once, including the shared attention count, until it's restored.
Snooze or permanently dismiss any alert. A snoozed flag comes back if the problem gets worse, or after a set window passes; a muted one stays quiet until you restore it, so you decide what you're actively being told about.
A quiet, dormant readiness to flag when it's time to re-analyze. The underlying engine that will eventually detect meaningful drift in your numbers and suggest a fresh analysis is built into the product today, running quietly while it waits for enough history to accrue on a given process.

What gets done

Whether the phased plan is actually being followed

Progress

Metrics asks whether the numbers are moving. Progress asks whether the work is actually getting done, tracked against the exact phased plan the analysis produced, not a separate checklist someone has to keep in sync by hand.

Grouped by the plan the analysis already wrote. Recommendations appear under the same implementation phases from the report's own roadmap, and anything not yet scheduled still shows, clearly labeled "Not yet scheduled," rather than left off.
Five fields, updated inline. Status, percent complete, a start date, an end date, and a named owner, each editable directly with no separate save step.
An honest status for work that's simply not happening yet. Alongside Not started, In progress, and Done, a change can be marked Deferred, deliberately distinct from "won't do," since almost nothing is truly never going to happen, and deferred work doesn't count against the completion total.
Quick wins tracked as their own work, not folded in. A quick win can be linked to the recommendation it relates to, but it carries its own independent status and owner, so closing one out never silently closes the other.
Every item shows exactly what value it supports. Each recommendation names the specific value pool it feeds and that pool's dollar figure, and if it contributes to a pool without being the one counted toward the total, that's shown too, so nothing is claimed twice.