Your ISO 45001 system,
drafted and running.
Describe the company once. SafetrexIQ drafts the management system, twenty policies and a scored risk register, then runs every incident, RIDDOR duty and corrective action on one record. You stay the competent person.
NO CARD REQUIRED · HOSTED IN FRANCE · AI OFF BY CHOICE
Illustrative example, fictional company. A profile is described once, SafetrexIQ drafts the cited management system, then an incident runs through the RIDDOR check, a systems-based investigation and a corrective action, closed and then verified, on the same record.
A safety management system is everything a competent person needs to run the duty: the policies, the roles, the scored risks, the incident record and the proof that actions worked. SafetrexIQ drafts it from one profile and keeps it current, for good.
BUILT HANDS-ON WITH A CHARTERED UK HEALTH AND SAFETY PROFESSIONAL · IN LIVE PILOT WITH A UK LOGISTICS OPERATOR
Likelihood × Severity
Hierarchy of control
Animation: a hazard is scored 4 × 4, High, and escalates by rule; an engineering control is recorded in hierarchy order and the hazard is re-scored to Medium. Hover to pause, then try any control or point at a cell.
RISK REGISTER
Score it once. Watch it move.
A 5×5 matrix whose ratings and escalation come from fixed rules, not AI. Record a control in hierarchy order and re-score the hazard; High and Extreme ratings escalate by rule. The same incident type twice in 90 days is flagged.
INCIDENT · IN-0412
Slip on wet floor, Warehouse B
reported via QR, 07:42 · anonymous
- AOReviewer: A. Okafor, competent person
- RDRIDDOR check run
Animation: an anonymous slip report is logged, SafetrexIQ drafts a systems-based investigation for review, the corrective action is owned and dated, then closed, and a competent person records whether it worked.
INCIDENT LOOP
Every incident closes with a verified action.
A report comes in, SafetrexIQ drafts a systems-based investigation for your review: contributing factors, each with the control that may have failed. The corrective action is owned and dated, and once it closes a competent person records whether it worked. The same incident type twice in 90 days is flagged.
SafetrexIQ does not file the report or track the deadline. Deadlines and filing remain your duty.
Animation: the facts of an incident are logged, rules judge it Reportable with the regulation cited, and the duty to report stays with your competent person. Hover to pause, then pick any outcome.
RIDDOR, PLAINLY
Reportable or not, decided by rules.
RIDDOR makes certain events reportable to the HSE. SafetrexIQ gives a rules-based reportability verdict from the facts you log, with the regulation cited, and tells you what it cannot decide for you. Deadlines and filing remain your duty: it does not file the report.
Animation: a phone scans a QR code on the wall, the reporter fills in the form with no login, gets a status page, and the report lands on the dashboard for triage.
ANONYMOUS INTAKE
Anyone can report. Nobody needs an app.
A QR code on the wall opens a report form: no login, no install. The reporter gets a status page so the report never vanishes, and the record lands on the dashboard for the H&S team to triage.
Each field is classified before the request runs.
Names become placeholders. Detected health information stays on our own model. Detection uses word lists and rules, so it can miss wording they do not cover.
AI drafting and the assistant are off. Nothing is sent to OpenAI.
Decided by your organisation's AI mode · every request audited
Draft investigation for PERSON_A1
DRAFT · FOR REVIEW
Animation: each field of an incident record is classified, the name becomes a placeholder, detected health information stays on our own model, and the draft comes back with names restored on your side. Hover to pause, then pick a mode.
AI PRIVACY
Names are replaced before anything leaves our servers.
Records are stored in France. The question and record text of every AI request is scanned by our privacy gateway before it can leave our servers; fixed instructions, including your organisation's name, are sent as written. Names recorded in your workspace, and names written with a title, are replaced before anything leaves our servers. Detected health information stays on our own model. Detection uses word lists and rules, so it can miss wording they do not cover. Each organisation chooses its AI mode, from Standard to Off, set by an owner, admin or H&S manager.
Animation: the module directory fills in, a search for RIDDOR narrows it to one card, the card opens to show what it shares with Incidents and Corrective actions, then closing the action updates the incident on that same record.
PLATFORM · 24 MODULES
Every duty has a place to live.
One record, one set of roles, one audit trail. Incidents, investigations and corrective actions share one record, and nobody digs through a binder for the current version.
Retire the binder.
Print twenty policies. File them. Lose the current version. Someone asks for the risk assessment at 4pm. Or keep one record, approved once, current everywhere.
Approve a draft
Draft → reviewed → approved, with the approver on record.
- Work at heightdraft · HASAWA s.2
- Manual handlingdraft · MHOR 1992 reg.4
- Lone workingdraft · MHSWR reg.3
A reference for every report
No login. They get a reference code and a page showing any update your team posts.
Nothing moves on its own: the page shows exactly what your team has posted.
Audit-ready PDF
Any generated document as a PDF, with its approval state.
Overdue actions reach the H&S leads and the owner
A missed due date is pushed, not discovered at the audit.
- In appthe notification bell
- Emailwhen email is configured
Action CA-118 overdue → the H&S leads and M. Okafor, its owner. RIDDOR-reportable verdicts alert the same leads.
Ten roles, one organisation, enforced server-side
Roles are organisation-wide. Hover one to see what it can reach.
The management system, risks, incidents, actions, audits and reports.
Report SX-0413 · what the reporter sees
Updates your team posts, as they post them.
- Received, 07:42
PRICING
Priced per organisation, not per worry.
Three typical sizes of the same system, as a guide rather than a limit. We scope the plan on the call: sites, headcount and how much of the AI you want on.
SOLE TRADER
One site
Everything in the platform, typically for a single site.
Typical size: 1 site
- Drafted management system
- Incident loop and CAPA
- RIDDOR verdicts, regulation cited
- AI drafting
- OH / absence data separation
SME
The whole loop
Typically several sites, on the same platform.
Typical size: several sites
- Drafted management system
- Incident loop and CAPA
- RIDDOR verdicts, regulation cited
- AI drafting
- OH / absence data separation
ENTERPRISE
Scoped on a call
For larger or more complex organisations. Tell us what you need and we scope it with you.
Typical size: many sites
- Drafted management system
- Incident loop and CAPA
- RIDDOR verdicts, regulation cited
- AI drafting
- OH / absence data separation
- SSO and role mapping: on the roadmap
Prefer to look around first? Create your organisation. No card required.
No. It drafts a management system aligned to the standard and runs it. Certification follows an independent audit by an accredited body.
DEPLOYMENT
Your safety system,
drafted the same day.
Create your organisation, answer a short profile, and SafetrexIQ drafts the system for your review. You stay the competent person; it does the assembly and the follow-through.
Generated frameworks, policies and risk assessments are drafts for a competent person to review and approve. Nothing here confers ISO 45001 certification, which follows an independent audit by an accredited body. Jurisdiction support is United Kingdom first.