Opening a safety record: the polymorphic detail deck

What it is

Every incident, written policy, and insurance policy on your Safety section opens into the same detail deck. Verinode calls this "polymorphic" internally because one shared engine renders four different kinds of record (incident, certification, policy, insurance) with the same hero layout, the same tab structure, and the same peer-comparison panel, switching only the content inside each slot to match what kind of record you opened. You get one mental model for "how do I read a safety record" no matter which one you clicked.

This is not a compliance tracker that just lists dates. Verinode reads the incidents, policies, and insurance data already flowing in from your business, works out where you actually stand against a peer cohort and against hard regulatory thresholds, and puts a single recommendation at the top of every record: what to do, what you gain by doing it, and what it costs you to leave it alone. You decide what happens next.

Where to find it

Open Safety from the sidebar at /safety. The page loads straight into the Safety home, a row of tiles summarizing incidents, expiring coverage, and policy currency. Clicking a tile opens an overlay slider with tabs across the top: Findings, All, Standing, Exposure, and Benchmarks (each tab reads in plain language, not raw code names).

From inside that slider, clicking or tapping a specific record flips the card in place to reveal the detail deck described in this article, the same card, showing its back face. Closing it (Escape, or the back control) flips you back to the row you came from.

If you arrive through a link from the Feed or from a decision (for example, ?findingId=... or ?recordId=... in the URL), Verinode jumps you straight into the Findings tab or drills directly into that record, skipping the manual click.

The stance-first hero

The top of every record leads with a posture, not a date. Above the title sits a small eyebrow row: the kind label (Incident, Policy, or Insurance) next to a colored pill spelling out the record's current stance:

• Healthy (Deere green), keep the streak going, nothing needs attention.
• Drift (Hard Hat yellow), approaching a threshold, worth fixing before it becomes an exposure.
• Exposed (Ember red), past a soft threshold, closeable this week.
• Breaching (deep red), past a hard threshold that is already blocking or costing you, fix it now.
• Watching (neutral gray), not enough data on file to stake a posture yet. This is an honest placeholder, not a failure.

Below the title, the subtitle is the record's single "what to do" sentence (for example, "Close the corrective action today" or "No action, review window opens 60 days before the deadline"), the same sentence that also anchors the agent's read further down the page. Every stance is computed fresh from your data: an incident's corrective-action due date and severity, a policy's review date and acknowledgment status, an insurance policy's expiration date, auto-renew flag, and (for workers' comp) its EMR.

While the record's full detail is still loading, the hero paints instantly from data already in memory (the kind label and the title), so you are never staring at a blank screen. The moment the deck finishes loading, the stance pill and the KPI numbers on the right replace that placeholder.

A Quick Survey button sits in the top-right of the hero on every record. It opens a one-question survey you can fire off to a team member about that specific incident, policy, or insurance policy, useful for a quick "did you follow up on this" check without leaving the page.

The KPI numbers (per kind)

To the right of the title, a small block of numbers gives you the record's vital signs at a glance. What appears there depends on which kind of record you opened.

Incident

• Severity, one of first aid, recordable, lost time, or fatality.
• Recordable, Yes or No, whether this incident counts toward your OSHA recordable rate. Recordable incidents are flagged in a different color than non-recordable ones.
• Corrective Action, reads "Closed" once resolved, otherwise a countdown or overdue count against the due date you set ("12d to close", "3d overdue"), or "No due date" if none was ever set.
• Logged, the date the incident was recorded.

Policy

• Review, the policy's next scheduled review date.
• Version, the policy's version number.
• Acknowledged, how many of your active team members have acknowledged the policy out of your total active team, shown as a fraction and, below it, what percentage of the team that represents.
• Status, Draft, Active, Under Review, or Archived.

Insurance

• Expires, the policy's expiration date, with days-left or days-overdue called out beneath it.
• Coverage, the coverage limit, formatted (for example $2.0M).
• Premium, the annual premium.
• The fourth number switches by policy type: on a workers' comp policy it is EMR (your experience modification rate, to two decimal places); on any other policy type it is Deductible.

How the deck loads: one batch, then a quiet second wave

Opening a record fires two separate requests, timed so you never wait on the slower one before seeing the fast one.

1. 1The primary batch fires immediately. A single call bundled by record kind fetches everything the hero and the Details and Open Tips tabs need in one round trip: for an incident, that is the linked job, the reporter's name, how many other OSHA recordables you have logged in the last 90 days, and the most recent prior incident of the same type; for a policy, that is who on the team has acknowledged it and the policy's older versions; for insurance, that is your other active policies and, when relevant, your workers' comp EMR. This is what paints the hero, the Details tab, and the Open Tips tab.
2. 2Linked Findings load a moment later, deferred. Once the primary batch resolves, a second, separate call fetches any open Decisions tied to this specific record and fills in the Findings tab quietly in the background. This keeps the hero's paint fast; the Findings tab simply arrives a beat after everything else and its badge count updates once it's in.
3. 3The peer-comparison numbers load in parallel with step 1, scoped to National by default, and refetch automatically if you switch the scope.

If the deferred Findings request fails for any reason, the Findings tab just stays hidden rather than showing an error, the rest of the deck is unaffected.

The agent's read

Directly under the hero sits Verinode's always-visible recommendation card. It restates the "what to do" headline, then two supporting lines: what you gain by acting, and what happens if you don't. For an open recordable, for example, that reads as something close to: acting fast with a clean root-cause pass roughly halves your odds of a repeat in the next twelve months; leaving it open compounds against your EMR, which is a three-year rolling average.

How You Compare

Below the agent's read, a peer-comparison panel shows how your numbers stack up against other operators, using the same National / State scope switcher you'll find on other detail decks across the platform. Four metrics appear here:

• Incident rate (per 100,000 labor hours)
• Training compliance
• Days since last recordable
• Policy review currency

Each metric shows your value next to the peer cohort's value at the scope you've selected, and, where available, an authoritative research reference (labeled by its source, for example an OSHA national average). Switching scope from National to State refetches the peer side of the comparison; your own numbers don't move.

The panel header shows how many peer operators are behind the comparison at the selected scope. If a metric has no operator value at all, it simply doesn't appear in the panel. If every visible metric is missing a peer value at your current scope, the panel reads: "Not enough peers in this scope yet. Switch scopes above or check back as more operators contribute. Industry research values are shown as a fallback."

As an independent data trust, Verinode never sells any of this data to carriers, and every anonymized number that feeds this comparison also produces a return for the operator who contributed it. This panel only surfaces four of the metrics the Safety section tracks at the portfolio level; the section's own Benchmarks tab (in the cards slider) carries the full set, including near-miss ratio and corrective-action closure time.

The Details tab

This is where the record's own facts live, laid out differently for each kind.

Incident splits into two blocks. What Happened shows the incident description, the location if one was logged, anyone involved if that was captured, and the linked job (claim number and client) if this incident ties back to a specific job. Root Cause & Preventive Change shows the 5-Why root cause you've documented and the corrective action you've assigned, each with its own due date and closed-date once resolved. If no root cause has been recorded yet, that field explains that a 5-Why pass is the fastest way to prevent a repeat, and that the incident-reporter agent can walk you through one. If no corrective action has been assigned, it reads: "No corrective action documented. Assign one with a due date to close the loop." When there's a pattern worth flagging, a third block, Pattern Context, appears showing how many OSHA recordables you've logged in the last 90 days and, if one exists, the date and severity of the most recent prior incident of the same type.

Policy shows a Policy Overview (category, version, effective date, review due date), then, when you have an active team to measure against, an Acknowledgments block naming who has and hasn't signed off, and a Version History block listing earlier versions of the same policy with their effective dates and statuses.

Insurance shows a Coverage Overview (carrier, policy number, effective and expiration dates, whether auto-renew is on), with the EMR called out when the record is a workers' comp policy and a short note on which side of the 1.0 and 1.25 thresholds it sits. Below that, an Other Active Policies block lists your remaining active coverage for context, so you can see the whole insurance picture without leaving the record.

Open Tips

Any open signal Verinode has raised specifically about this record shows here, most recent first, up to twenty at a time. Each one shows a headline, the reasoning behind it, and, where there is one, a recommended next step. This tab only appears when there is at least one open tip tied to the record; once you act on a tip elsewhere in the product, it drops off this list. The tab's badge shows how many are currently open.

Findings

Any Decision linked to this specific record shows here: its title, and, when Verinode can size it, the monthly cost of leaving it unaddressed. This tab only appears once at least one linked Decision exists, and it loads a beat after the rest of the deck (see the loading section above). Opening a linked finding takes you into the decision workspace to work it.

Reading a record end to end

A typical read-through: check the stance pill first, it tells you whether this is a "keep doing what you're doing" record or one that needs this week. Check the subtitle for the one-line action. Scan the KPI numbers for the specific date or figure driving that stance. Read the agent's gain-versus-consequence pair to understand why it matters. Check How You Compare to see whether this is an isolated record or part of a pattern against your peers. Then open the Details tab for the underlying facts, Open Tips for anything Verinode is already flagging on it, and Findings for anything already queued to work.

Data sources

Related articles

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