Findings tab: equipment decisions grid

What the Findings tab is

Equipment has its own card slider, the overlay that opens when you click into the section from /equipment, and Findings is the first of its five tabs: Findings, All Equipment, Condition, Utilization, Benchmarks. Where All Equipment lists every unit you own, rent, lease, or finance and Condition tracks aging, calibration, and repair status, Findings is the tab that holds Verinode's actual work product for this section: the equipment decisions your agent has surfaced from purchase dates, condition logs, rental duration, and class density, laid out as a tile grid you can scan and click into one at a time.

Findings uses the exact same shared grid component, FindingsGrid, that every other section's Findings tab uses (jobs, vendors, clients). Nothing here is generic advice. Every tile traces back to a real pattern read from your own fleet: a unit aging past its expected life, a rental running long enough that buying would be cheaper, a class with no backup unit. Verinode surfaces the pattern and lays out the recommendation. You decide whether to act, park it, or ignore it.

Where to find it

Open Equipment from the sidebar (/equipment), then click into any unit, class group, or metric to open the card slider. The slider's tab strip runs Findings · All Equipment · Condition · Utilization · Benchmarks, in that fixed order, and Findings is selected by default. Decisions load as soon as the slider opens, in parallel with the open equipment tips that populate the Condition tab, so neither tab waits on the other.

The grid

At the top of the tab, a small uppercase label counts how many decisions are waiting: "N decisions to review" ("1 decision to review" when there is exactly one). Below it, the tiles lay out in a responsive grid, one column on narrow screens, up to three across on wide ones.

Empty state. If Verinode has nothing to surface for this section yet, the tab reads exactly: "No decisions for equipment right now. As your agent finds patterns, they'll appear here." This is not a broken screen. It means no equipment pattern (an aging unit, a long-running rental, a single point of failure in a class) has cleared the bar to surface yet, not that Verinode has stopped looking.

Reading a tile

Each tile is built the same way every Findings gallery on the platform is built, so once you know how to read one here, you know how to read them everywhere:

• The dollar figure, when there is one. A large bold number at the top of the tile, formatted the same way it is across the platform ($450, $12k, $1.4M), with a small period label beside it ("/mo" for a monthly figure, "/yr" for an annualized one, or "one-time" for a pool that shouldn't be read as recurring). This is the impact Verinode is putting a number on: a rent-vs-buy break-even, a repair-or-replace cost, a redundancy gap tied to job-cycle drag.
• The title underneath. A short, readable line naming the unit or class and what's going on with it, for example a specific unit past its expected life or a class of equipment with no backup on hand. This is never a raw database value, it's always written as a sentence.
• No dollar figure yet. Some findings can't be cleanly priced yet. On those tiles, a small risk-kind and urgency chip (for example "Business risk · Act soon") sits above a bolded one-line consequence statement instead of a dollar figure, with the full title underneath in smaller type.
• The status pill, top right. A small colored dot plus label. On the Equipment Findings tab this reads Pending, in Hard Hat Yellow, since a freshly surfaced decision hasn't been acted on yet.
• The action row at the bottom. Three buttons: Act, Not now, Ignore. On this tab the row renders visibly but grayed out, it's not wired for one-tap action directly from the tile grid here. Clicking anywhere on the tile itself (not one of the three buttons) is what opens the decision.
• The "Acted" or "Done" stamp. If a decision you've already acted on or resolved is still showing here, the tile tints green and carries a rotated rubber-stamp watermark reading ACTED or DONE. Clicking that tile skips the drill-in entirely and takes you straight to the decision's full workspace at /decisions/[id], since there's nothing left to review on this tab, the action plan is where the work continues.

Clicking into a decision

Click any tile (other than an already-stamped one) and the slider drills one level deeper: instead of the grid, you're looking at a single decision, full detail, filling the same overlay. A breadcrumb at the top, "← All decisions," in copper, takes you back to the grid.

Navigating between decisions. Once you're drilled in, you don't need to back out to see the next one. Swipe left or right (trackpad or touch), use the Left/Right arrow keys, or click the floating side arrows the overlay provides, and the deck pages through your open equipment decisions one at a time. A small line at the bottom of each decision reminds you: "← / → or swipe to navigate decisions · Esc to go back." Press Esc at any point to close the drill-in and return to the grid.

What the decision detail shows

The drilled-in view is the full decision card, the same one used across the Feed and the /decisions workspace, rendered flat here (no extra card frame, since the overlay already provides the surface).

Header row. A small dot colored to the decision's trajectory (pulsing red for declining, green for improving, yellow for stable), a status word (Declining, Needs context, or plain Decision), the unit or class name, and, on the far right, a confidence pill (high / medium / low), colored green, yellow, or gray.

The impact number. When Verinode has a dollar figure, it's the largest thing on the card: the annualized or monthly amount at stake, with a plain-language suffix ("/yr at stake," "/mo at stake," or "one-time at stake"). Directly under it, a small tag reads Calculated (green, when the figure is derived directly from your own equipment data, purchase price, rental days, condition log) or Estimated (yellow, when it's a working assumption). When Verinode knows more about where the estimate came from, a second tag chains on: "· Peer cohort" (built from how similar operators run comparable fleets) or "· Industry baseline" (built from published restoration-equipment benchmarks rather than a live peer comparison). Beside the tag, a short sentence spells out the calculation in plain language when one is available.

No dollar figure. For findings that can't be priced cleanly yet, this slot instead shows a risk-kind and urgency chip (the same pairing as the tile) above a bolded one-line consequence statement.

The title. The same headline sentence you saw on the tile, now given its own line.

Lifecycle dots. A four-stage tracker, Flagged → Planned → Acting → Resolved, the same pattern jobs use for Assigned/Started/Billed/Paid. On a fresh decision this sits at Flagged, so you can tell at a glance the decision hasn't been committed to yet.

The recommendation, or the honest gap. One of two tiles, never both:

• Recommended action (copper border): the concrete next step Verinode is proposing, most commonly buy this rental out, retire and replace this unit, or add a backup to this class, plus a line reading "If you don't: [consequence]" and, when confidence isn't high, a "Refine this estimate" note naming what data would sharpen the number, exactly what forwarding a purchase receipt or answering the agent's follow-up would fix.
• Needs your context (dashed border): when there isn't enough to recommend a specific action yet, this tile states the pattern plainly and tells you that clicking Act hands you to the agent, who will ask a couple of quick questions to shape a real recommendation instead of making you fill out a form.

What peers did. Below the recommendation, a short block (hidden entirely when Verinode doesn't have cohort coverage for this particular signal yet, no placeholder, no "insufficient data" message) reads how many operators facing a comparable equipment pattern in the last 90 days acted on it, and, of those, how many saw the underlying metric improve afterward. This is read qualitatively, as a share, never as a specific peer headcount, and it's the moat made visible: proof that acting on this kind of signal has actually worked for others in your position, not a guess.

The buttons. Act (primary, copper) hands the decision to your agent and takes you to its full workspace at /decisions/[id], where the plan gets built. Not now opens a short reason picker, Too busy, Need more info, Not convinced, or Other, and parks the decision (a confirmation reads "Parked, this will resurface later"). Why?, shown only when Verinode has proof to show, expands a drawer underneath with the specific numbers behind the finding (your own value, the peer average, and the research standard, when each exists) and, when known, the root cause chain that led to this decision.

What happens after you act

Clicking Act doesn't do the work invisibly. It materializes the underlying decision as a real, trackable item and routes you to /decisions/[id], the same workspace every decision across the platform shares, where your agent drafts a concrete plan you can review, edit, and check off step by step. See The decision workspace for how that full lifecycle works, from a fresh Act click through to a resolved outcome.

How Findings relates to the rest of Equipment

Findings is where Equipment's work product lives; the other four tabs are where the underlying data lives. A decision flagging a unit past 80% of its expected life traces back to the aging list on Condition. A decision flagging a rental running long enough to buy out traces back to the rentals list on Condition and the ownership split on Utilization. A decision flagging a class with no backup unit traces back to the class density rollup on Utilization. A decision comparing what your fleet costs against what peers run traces back to Benchmarks. Findings is the tab that turns all of that raw fleet data into something you can act on, one decision at a time, instead of leaving you to spot the pattern yourself across four separate tabs.

Equipment also has its own lighter-weight surface for open patterns: the Condition tab shows a running list of active equipment tips generated by the detector, each with its own Act / Not now / Ignore row, and acting on a tip opens the same kind of pre-drafted decision workspace that a Findings tile does. Findings is the fuller gallery; Condition tips are the same intelligence surfaced in-line where the underlying unit lives.

Best-practice example

Say the Findings tab reads "2 decisions to review." The first tile carries a dollar figure of roughly $3k a year and a title naming a specific unit whose age has crossed into the aging window. Click it. The drilled-in view shows an Estimated · Industry baseline tag on the number, meaning the figure is built from published equipment-lifespan and repair-cost data rather than a live peer comparison. The recommendation tile suggests planning a replacement before the unit's reliability risk compounds further, with a "What peers did" block showing that most operators who acted on a similar aging-equipment finding saw the underlying repair or callback pattern improve within a few months. Click Why? to see the unit's age against its expected lifespan before you commit, then Act to send it to your agent, who drafts the replacement plan in the Decisions workspace.

Data sources

Related articles

• The decision workspace
• Benchmarks overview
• Reading a benchmark
• How benchmarks work
• Understanding your margin