Sentinel AI · Site Safety, Grounded

Speak a hazard. It's logged, and escalated.

Sentinel AI turns your own safety manual into an always-on WhatsApp reference — and turns a spoken hazard report into a structured record, automatically routed to your safety lead. No new app to install.

Answers only from your approved manual — escalates to a human when it isn't sure

WhatsApp · Live
Worker"Exposed rebar near the east stairwell, level 3 — no caps, high traffic."
Sentinel AILogged as: Struck-by/Impalement risk · Level 3, East Stairwell · Priority: Immediate. Routed to Site Safety Manager.
Worker"What's the barricade spec for an open trench edge?"
Sentinel AIPer your Site Safety Manual §4.2: minimum 1.1m barrier, 2m setback, warning signage at 10m intervals.
35
construction fatalities in Great Britain, 2024/25 — the highest of any sector
~50%
of construction fatalities are falls from height, the leading single cause
4.8x
construction's fatal injury rate vs. the all-industry average
~50,000
non-fatal injuries recorded in construction, 2024/25

Source: Health and Safety Executive (HSE), Great Britain, 2024/25 official statistics.

The Problem

Hazards get mentioned, not logged.

Site teams already know the risks. What gets in the way is stopping work, finding the right form, and typing a report one-handed while wearing gloves — so the hazard gets a mention at the toolbox talk instead of a record, and the pattern never gets tracked.

  • Paper forms and clipboards get lost or filed too late to act on
  • A 100-page safety manual isn't something anyone reads mid-shift
  • Safety officers can't be everywhere, on every site, at once
  • Generic reporting apps mean a new login and a new habit
The Sentinel AI Fix

Your safety manual, always on. Reports, always logged.

Workers message a question or a hazard over WhatsApp — the tool they already use. Sentinel AI answers strictly from your own approved manual, or turns a spoken hazard into a structured, routed report in seconds.

  • Hands-free reporting — no app to install, no new login
  • Answers grounded only in your site's approved manual
  • Automatic hazard classification, location, and priority tagging
  • Escalates to your Safety Manager when it's uncertain — never guesses
What's Included

A digital safety reference your team already knows how to use

Built for site managers and safety officers — not developers.

💬

WhatsApp-Based Access

No app to download. If a worker can send a WhatsApp message, they can use Sentinel AI.

🎙

Voice-to-Report

A spoken hazard becomes a structured, tagged, routed report — automatically, in seconds.

📘

Grounded to Your Manual

Answers only come from your own approved safety documentation. If it's not in the manual, Sentinel AI says so and escalates.

🔺

Automatic Escalation

Time-sensitive or uncertain reports route straight to your Safety Manager — no report sits unread.

🗣

Digital Toolbox Talks

Daily briefings logged digitally, replacing paper sign-in sheets and helping demonstrate compliance.

🔒

Private, Per-Client Knowledge Base

Enterprise-grade encryption. Your manual and reports stay yours — never used to train public models.

How It Works

From spoken hazard to routed report

Nothing new to install — Sentinel AI runs over the messaging app your team already has open.

1

Message Sentinel AI

A worker sends a voice note or text — a question, or a hazard they've just spotted.

2

It answers, or it structures the report

Questions get answered strictly from your approved manual. Hazards get classified, tagged, and prioritised automatically.

3

The right person gets alerted

Your Safety Manager receives the routed report instantly — and every report is logged for audit.

Pricing

Flat-rate, not per-seat

Confirmed pricing — no usage roulette, no per-worker fees.

Starter
£249 / mo
+ £350 setup · single site
  • WhatsApp safety Q&A, grounded in your manual
  • Digital Toolbox Talk sign-off
  • Hazard escalation to 1 named Safety Officer
Get Started
Most Popular
Growth
£399 / mo
+ £650 setup
  • Everything in Starter
  • Up to 3 sites
  • Multi-person escalation routing
  • Monthly compliance report
Get Started
Scale
from £750 / mo
+ from £1,200 setup
  • Everything in Growth
  • Unlimited sites
  • Dedicated knowledge-base management
  • Quarterly review
Talk to Sales
Before You Ask

The questions every site manager asks first

Is our data safe?

Enterprise-grade encryption, a private knowledge base per client, and your data is never used to train public models.

What if it gives the wrong safety advice?

It doesn't guess. Sentinel AI only answers from your own approved manual — anything outside that scope is escalated to a human, not invented.

Our workers aren't tech-savvy — will they actually use it?

It runs over WhatsApp or your existing messaging tool. No new app, no training curve, no separate login.

We already have a Safety Officer — isn't this redundant?

Sentinel AI scales your Safety Officer's expertise 24/7 rather than replacing them — putting their knowledge in every worker's pocket, and freeing them for higher-level risk work.

Looking Ahead

We're exploring a wearable hazard-detection device as a future direction for Sentinel AI. It isn't available today, and nothing above depends on it — this is a long-term vision, not a current feature.

Get Started

See Sentinel AI on your own jobsite

Book a short walkthrough — bring one real safety question or hazard scenario, and we'll show you the full path from message to routed report.

Confirmed pricing above · setup is fast, but Sentinel AI's live backend is still being connected — book a walkthrough to discuss real timelines for your site.