Capture early signals before they escalate.
Sentinel is PsychProof's enterprise continuous monitoring engine — delivering scheduled and event-triggered psychosocial check-ins via Microsoft Teams webhooks, with k-anonymity enforcement and automatic evidence trail generation at scale.
For organisations that need the Change Planner or core platform first, start at psychproof.com.au
Distributed observations.
Centralised intelligence.
Raw signals bypass local line management completely, flowing directly into a cryptographically secured central vault.
Just a new contact for your team.
There is no app to download, no new login to remember, and no training required for your workforce.
Sentinel meets your team where they already are. Slack channel for signal capture, pinned intake link for anonymous reporting, email pulses for periodic check-ins, web for everything else. No new app to install. No new login to remember. Thirty seconds to log an observation, in the flow of the existing workday.
A micro-observation
Standard surveys wait for employees to complain. We capture the subtle operational realities that managers observe every day.
It could be a team working sustained overtime, subtle interpersonal friction, or visible pressure during a roster change. These are the lead indicators of psychosocial risk, captured long before they escalate into formal HR grievances.
Workers see things their supervisors don't.
Most platforms aggregate one channel. PsychProof compares them — and shows you where the reality on the ground doesn't match what's coming up your reporting line.
Context, the moment a signal arrives.
Signals don't arrive in a vacuum. PsychProof surfaces what your WHS team needs to know — automatically, inline, before they decide what to do.
Pattern
This is the Nth signal in M weeks for this hazard at this site. The platform tracks frequency and surfaces it inline.
Divergence
Workers are reporting it. Are your supervisors? The platform compares the two channels and shows the gap.
Jurisdiction
What your state's regulation actually requires for this hazard class — surfaced as the manager triages.
Reasoning Engine Active
Hierarchy of Control Guidance
Elimination recommended. Review rostering overlap to remove standalone shift handover pressure.
Three layers of proactive signal capture.
Sentinel doesn't just wait for annual surveys. It actively seeks out risk indicators where they are most likely to occur.
1. Scheduled Cadence
The baseline rhythm. A configurable recurring check-in (default: 14 days) sent to all employees. It ensures continuity of the evidence trail and establishes organisational rhythm without waiting for a crisis.
2. Event-Triggered
High-signal prompts fired by operational conditions known to elevate risk: restructures, key person departures, sustained overtime, or return-to-work events. Sent exactly when observation is needed most.
3. Silence Signals
Non-response is data. Sentinel tracks response patterns and automatically escalates when an employee misses consecutive check-ins in a high-pressure environment. The system cannot be gamed by silence.
Five minutes to set up. Thirty seconds per employee.
A live evidence trail from day one.
Targeted or Routine
Whether triggered by an operational event (like a restructure) or running on a fortnightly cadence, Sentinel knows when observation is needed most.
Inside Microsoft Teams
Managers receive an Adaptive Card right in Teams. One question, optional tags. Completed in 30 seconds without leaving their workflow. No new login required.
Live Defensible Record
Responses are timestamped, hashed, and logged into the compliance trail. Patterns escalate automatically to HSEQ before they become incidents.
From SafeWork Notice to Proven Compliance in 90 Days.
A mid-size Australian mining operator received an improvement notice after two consecutive stress-leave events on the same crew. They needed to prove active monitoring, fast. Here is exactly how Sentinel closed the loop.
Targeted check-in deployed
HSEQ imported 18 supervisors across two sites. A specific trigger was sent to the Drill & Blast crew leader regarding a recent roster change, alongside the standard 14-day cadence for all managers.
Critical fatigue signal flagged
Drill & Blast supervisor logged an observation via Teams: "Two blokes told me separately they're not sleeping well. New roster rotation means some guys are doing longer drives at the end of a swing."
SafeWork notice closed
During the follow-up inspection, HSEQ presented the Sentinel dashboard showing 126 timestamped responses, an 83% response rate, and documented control actions for the fatigue signal.
Privacy by design.
The condition for honest participation.
K-Anonymity Protection
Responses aggregate into team patterns. Minimum group thresholds (k-anonymity) prevent isolation and guarantee individual attribution is mathematically impossible.
Access-Controlled Evidence
The forensic trail is accessible only to compliance officers, never to line management.
What employees see on submission:
"Your observation has been logged. It is not visible to your line manager. It is not a report on your performance or your team. It goes to your HSEQ team, who will determine whether any support or action is needed. If your signal is escalated, you will be notified. Thank you for observing."
Flagging a signal is leadership, not a liability.
Employees and supervisors sometimes hesitate to report what they're observing. The concern is understandable: if I say my team is under pressure, does that reflect on my performance? Will HSEQ investigate us?
Sentinel is designed specifically to remove that barrier.
You have made a professional observation and put it on the record. That is exactly what your WHS obligations expect of you — and it is the evidence that shows you were paying attention.
A flagged signal is not a performance report on you or your team. It is not the start of an investigation into your management. It does not go to your line manager. It goes to your HSEQ team — who decide what support or action is appropriate.
If your signal is escalated, you will be told it is being reviewed — nothing more. When something is acted on, you will be told that too. You will not be identified as the source in any leadership or board-level view.
Here is the harder truth: if something does go wrong later and you never raised any signals, the question will be why didn't anyone flag this? Sentinel gives you the answer — because you did.
One system. Three different conversations.
A live signal map of what your employees are experiencing
The dashboard shows patterns across teams, sites, and time. Cluster alerts surface automatically when similar signals appear from the same location within a single cycle.
Documented proof that monitoring is occurring
The officer view shows active monitoring status, escalation rates, and evidence trail integrity. Section 27 due diligence, on the record, ready for board reporting or regulator inquiry.
Thirty seconds, once a fortnight
A check-in arrives in your workflow. One question. No login. Your observation is logged and confidential — it does not go to your line manager and it is not a performance measure. Flagging a concern is the evidence that you noticed and acted professionally. You find out when something you raised has been acted on.
Deploy targeted triggers.
Monitor the engagement gap.
Targeted Event Triggers
Don't rely on generic cadence alone. When an operational event occurs—like a restructure, a key departure, or sustained overtime—HSEQ can deploy context-specific questions targeted precisely to the affected crews.
Deploy Event Trigger
Targeted contextual check-in
✓ 3 Managers selected
"Your crew has recently had a significant roster change. Are there signs of adjustment difficulty, increased pressure, or uncertainty in the new configuration worth noting?"
Silence as Data
The system cannot be gamed by employees simply ignoring it. Sentinel tracks response patterns and auto-generates Engagement Gap alerts when an employee misses consecutive check-ins, enforcing the Duty of Care.
Engagement Gap Alert
Silence threshold exceeded
System Note: Non-response in high-pressure environments is treated as a potential signal. HSEQ manual follow-up is required to close this loop.
Every response. Timestamped. Hashed. Tamper-evident.
Sentinel doesn't just collect data; it creates a legally defensible record. Built continuously, in the background, from the moment Sentinel goes live.
"This is the first time I've had visibility into what employees are actually observing and experiencing at site level."
A resources company with two remote sites, a new site manager, and two stress leave events from the same crew in three months. No formal psychosocial monitoring in place. SafeWork improvement notice received.
126 timestamped check-in records. Two fatigue and sleep disturbance signals identified and escalated before a third stress leave event occurred. SafeWork improvement notice closed. First board-level visibility into psychosocial conditions across both sites.