What makes inviol different: the coaching-first approach to safety

When safety teams evaluate AI monitoring platforms, the pitch they usually hear sounds something like this: "Our system detects unsafe behaviours in real time and sends alerts so you can act before an incident occurs."

That is not a bad pitch. Real-time detection is genuinely useful, and most of the platforms that deliver it do so well. But there is a question that pitch leaves unanswered: what happens after the alert?

Because alerts, on their own, do not change behaviour. They generate a notification. Someone has to receive it, review it, decide what to do with it, have a conversation with the worker involved, follow up, and confirm that something has actually changed. In a busy facility running 24/7, with shift supervisors managing multiple priorities, that follow-through is exactly where most safety programmes start to break down.

This is the problem inviol was designed to solve. Not just detection — detection connected to a coaching process that actually produces behaviour change. It sounds like a small distinction. In practice, it is a significant one.

Most platforms are built around the alert. inviol is built around what comes after it.

The majority of computer vision AI safety platforms are, at their core, monitoring and alerting systems. They watch your cameras, detect defined event types, and send a notification when something is flagged. That is genuinely useful for situational awareness and compliance documentation. It is less useful for changing the day-to-day behaviours that determine whether incidents happen.

inviol's design philosophy starts from a different place. The goal is not to produce more alerts. It is to produce better coaching conversations — and to make those conversations easier for supervisors to have, more specific in content, and more consistent in follow-through.

The distinction matters because safety culture is not built through notifications. It is built through the quality and consistency of the interactions between supervisors and their teams. Every time a supervisor sits down with a worker to discuss what happened in a specific situation, why it matters, and what should happen differently next time, that is a small deposit into the safety culture account. Consistent, well-grounded coaching is what turns a detection system into lasting behaviour change.

What coaching-first looks like on the ground

In practical terms, here is how the coaching-first approach works in a typical inviol deployment.

When inviol detects a safety event — a worker entering a forklift exclusion zone, a vehicle approaching without adequate clearance, a PPE gap during a high-risk task — it captures a short clip of the event, with faces and people automatically blurred so that what the supervisor sees is the behaviour and the conditions, not a person to blame.

That clip becomes the starting point for a coaching conversation. Instead of a supervisor reconstructing what happened from memory (or not having the conversation at all because there is nothing concrete to show), they have an objective record of a specific event, in their own facility, that they can sit down and review with a worker or a team. The conversation moves from abstract ("you need to be more careful in the exclusion zone") to specific and grounded ("here is what happened on Tuesday at 6.15am, let us talk about what was going on and what we do differently next time").

This is a meaningfully different kind of conversation. Research on safety culture consistently shows that coaching grounded in real, specific events is more effective than generic reminders or policy reviews. Workers respond differently when the discussion is about something they can see and recognise, rather than a theoretical scenario or a rule being recited.

Toolbox talks, automated — with your own footage

One of the features safety managers tend to find most practically useful is inviol's automated toolbox talk capability. Rather than constructing a safety briefing from industry examples or generic materials, inviol can generate a site-specific coaching session from the events your own facility has produced. The footage is real. The scenarios are from your operation. The risk is something your team has actually encountered, not a hypothetical.

This changes the dynamic in a toolbox talk in ways that matter. Workers are more engaged when the example is something they recognise. The discussion is more specific. And the message is clearer: this is about how we work here, not a generic reminder from head office.

For facilities with mixed-language workforces — a common reality in warehousing, logistics, and manufacturing across New Zealand and Australia — visual coaching material drawn from actual site footage also significantly reduces the language barrier in safety communication. A short clip does not require translation. The risk is visible. That is a meaningful advantage for any organisation running on seasonal or agency labour with diverse language backgrounds.

Privacy built into the design

One of the most consistent concerns safety managers raise when evaluating AI monitoring platforms is worker privacy. How footage is stored, who can access it, whether workers feel surveilled, and whether the system creates more anxiety than it reduces.

inviol's response to this starts at the hardware level. The AI processing happens on-site — 99% of footage never leaves your facility. Only the short event clips (already face-blurred) move off-site, along with the event metadata needed to support reporting and coaching workflows. There is no cloud storage of raw surveillance footage.

The face and people blurring is automatic and non-optional — it runs on every clip, every time. This is a deliberate design choice that reflects the coaching-first philosophy: the point is to discuss the behaviour and the conditions, not to identify an individual and assign blame. That framing makes workers more receptive to coaching conversations, and it removes the adversarial dynamic that can undermine trust when monitoring technology is introduced.

inviol is SOC2 and ISO27001 certified with full GDPR compliance, and the trust centre at trust.inviol.com provides full transparency on data handling. For enterprise customers with specific IT security requirements, inviol's team can walk through the architecture in detail — something that matters increasingly as industrial operators face scrutiny from their own enterprise customers over supply chain data governance.

Existing cameras. No infrastructure overhaul.

Another practical differentiator that comes up consistently in buying decisions is the question of infrastructure. Some safety technology requires significant new hardware installation — sensors on vehicles, new camera networks, major IT integration projects. inviol does not.

In almost all deployments, inviol works with the CCTV infrastructure already in place. It connects via standard RTSP feeds and is compatible with the major camera and NVR brands already running in most facilities: Hikvision, Dahua, Axis, Hanwha, Avigilon, Bosch, and others. The hardware inviol provides — a small on-site AI unit — sits alongside the existing recorder and requires no modification to current equipment or significant downtime to install.

This matters for two reasons. The obvious one is cost: not having to replace your camera network before you can deploy is a real commercial advantage, particularly for multi-site operations where infrastructure standardisation across facilities is expensive and slow. The less obvious reason is speed to value. A deployment that can go live in days rather than months means you start getting useful data — and supporting coaching conversations — much faster.

What the data shows

The proof of a coaching-first approach is in the outcomes it produces, and inviol's customer data is clear.

Customers see an average risk reduction of up to 67%, with up to a 42% reduction in incidents over three years and up to a 61% reduction in machine-on-plant incidents — one of the most serious and costly event categories in warehouse and logistics environments. These are not short-term spikes following initial deployment. They are sustained reductions, produced by a consistent cadence of coaching conversations over time.

The difference between a facility that uses inviol for monitoring and one that uses it for coaching is visible in these numbers. Detection alone tends to produce initial improvements as awareness increases, then plateau as workers adapt to the presence of monitoring without the sustained follow-through of coaching. Coaching-first approaches continue to improve because they are building something — a safety culture, a set of behavioural norms, a shared understanding of why the rules exist — rather than just enforcing compliance.

NZ Post's General Manager of Safety and Wellbeing, Rex Middelbeek, articulated it directly: "Positive coaching equals a reduction in risk." That is the operating principle inviol is built around.

Who inviol is for

inviol is not the right fit for every organisation. To be direct about it: if you are primarily looking for a compliance documentation system, or if you are operating a site where the main safety challenge is capturing incident records rather than changing daily behaviour, there are simpler tools that may suit you better.

inviol is designed for organisations that are serious about improving safety culture — not just demonstrating compliance — and that recognise the coaching conversation as the mechanism through which that improvement actually happens. It is particularly well suited to high-risk industrial environments: warehouses, logistics and freight facilities, manufacturing, cold storage, ports and airports, construction materials yards, and processing operations where vehicle-pedestrian interaction and machine-on-person risk are the primary concern.

It is also a good fit for operations that have already tried purely alert-based systems and found that alert volume does not translate into behaviour change. That is a common experience, and it usually reflects the same underlying issue: detection without a structured coaching process does not change how people work.

If that sounds familiar, book a demo and we will show you what the coaching-first approach looks like in practice — using examples from facilities that operate in the same kind of environment as yours.

Frequently Asked Questions

What makes inviol different from other AI safety monitoring platforms?

Most AI safety platforms are built around alerts — they detect events and notify someone. inviol is built around the coaching conversation that follows detection. The platform connects detected safety events to structured coaching workflows, automated toolbox talk materials, and documented follow-through. This produces sustained behaviour change rather than short-term awareness spikes.

How does inviol protect worker privacy?

inviol processes footage on-site, meaning 99% of footage never leaves your facility. Only short event clips are used for coaching, and these are automatically face and people blurred before any review or use. The platform is SOC2 and ISO27001 certified with full GDPR compliance. The design is deliberate: coaching conversations focus on behaviour and conditions, not individual identification.

Do I need to replace my existing cameras to use inviol?

No. In almost all deployments, inviol works with your existing CCTV infrastructure, connecting via standard RTSP feeds. It is compatible with the major camera and NVR brands common in industrial facilities. inviol provides a small on-site AI processing unit at no additional hardware cost that sits alongside your existing recorder.

What kind of results do inviol customers typically see?

Customers see an average risk reduction of up to 67%, with up to a 42% reduction in incidents over three years and up to a 61% reduction in machine-on-plant incidents. These are sustained results produced by a consistent coaching cadence, not short-term improvements that plateau as novelty wears off.

What types of facilities is inviol best suited to?

inviol is designed for high-risk industrial environments where vehicle-pedestrian interaction and machine-on-person risk are primary concerns — including warehouses, logistics and freight facilities, manufacturing, cold storage, ports, airports, construction materials yards, and processing operations. It is particularly valuable for organisations that are focused on improving safety culture, not just documenting compliance.