Your ERP runs the business. We run the risk.
RiskSafetyAI sits alongside SAP, Dynamics, IFS, SYSPRO, Ellipse and Maximo — consuming the master data you already maintain and handing back the incident, exposure and compliance data they were never built to capture.
“So you’re an ERP?”
The honest answer, upfrontNo — though we look like one in places, because you cannot manage safety without knowing who works where, on what equipment, under whose supervision. Your ERP stays the source of truth. We read from it.
What we own is the part your ERP was never designed for: the hazard identified at the face at 04:00, the permit to work that expires in forty minutes, the contractor whose medical certificate lapsed yesterday, the Section 54 exposure nobody has quantified yet.
Read the full reasoning
We get asked this in almost every first meeting, and it’s a fair question. RiskSafetyAI holds employees, contractors, sites, cost centres and equipment. It issues documents, tracks approvals and reports on cost. From a distance, that looks like enterprise software doing enterprise things.
But we don’t own any of those records. We consume them. Your ERP remains the single source of truth for people, finance, procurement and assets — and RiskSafetyAI reads from it rather than competing with it.
Your ERP keeps its job. So do we.
Every layer keeps its job. We run the field half of each one — and hand the record back.
Mining operations run four distinct software layers. They interoperate, but they are not substitutes for one another — and the failure mode we see most often is a company trying to force one layer to do another layer's job. Running the field half of each layer is the opposite of that failure mode: ownership never moves. What each layer gains is a field presence it never had — the part of its work that today happens on paper between a work order being issued and being closed.
What we run here- Risk assessments
- Incidents & investigations
- Permits to work
- Inspections & audits
- Training & competency currency
- Contractor compliance
- Environmental & legal register
- Regulatory submission
RiskSafetyAI operates here. Our peer group is IsoMetrix, Cority, Enablon, Intelex and Sphera — not SAP. We compete there from a different starting point: offline capture at the working place, AI assessment on submission, and the platform configured in whichever languages your crews speak.
Start read-only. Unlock the rest at your pace.
Three phases. Each one stands on its own, delivers value before the next begins, and is a decision you make separately.
Read
Your ERP feeds us. Nothing changes in it.
- Master data flows in — people, contractors, assets, work orders, competencies, documents.
- Read-only, so the risk is effectively none. Every process you run today keeps running.
- No duplicate registers to maintain, and approvals follow your real reporting lines.
Execute
Chosen work moves to the field. One capability at a time.
- Your ERP keeps planning, scheduling and ownership of the record.
- We run the point of work — offline, with checklists, photos, permits and signatures.
- You pick each capability separately. You can take one and stop.
Two-way
The evidence flows back.
- Completion, hours, materials, findings, condition and cost return to the ERP — as part of your phase-three build.
- Maintenance planning starts responding to observed condition, not just calendar intervals.
- One complete history per asset and per job, spanning both systems.
Phase one is where every client starts — and on its own it delivers most of the value.
Both systems can do the job. You choose who runs it.
Your ERP was built for planning, cost and the enterprise record. We were built for the point of work. Where those overlap, nothing is required — each capability is a separate choice, taken at your pace.
Who runs the work order?
Your ERP plans and costs it. We can run it in the field.
How that works
A work order raised in your ERP appears on the crew's phone. They execute against a checklist, capture photographs, attach the permit to work if one is required, and sign it off — offline if they are somewhere without coverage.
Hours worked and materials used are captured as part of that, along with the evidence, and all of it returns to the record. We cover the part that currently happens on paper between a work order being issued and being closed.
Who decides when maintenance happens?
Your ERP schedules from intervals. We can raise work from observed reality.
How that works
Recurring work generates automatically — daily, weekly or monthly, assigned to a person or fanned out across a department. Beyond that, work can be raised by condition: an inspection returns a poor result, a check falls overdue, an asset's criticality passes a threshold.
Your ERP schedules maintenance from planned intervals; we schedule it from what your crews are actually recording in the field. Those are complementary inputs — the choice is whether the field evidence starts driving the maintenance plan.
Who owns contractors?
Your ERP owns the commercial relationship. We own permission to work.
How that works
Induction completion, document currency, insurance and certification expiry, site access — a contractor whose safety documentation lapses can be blocked from site automatically — enforcement level is your choice — without touching their commercial standing in your ERP.
We also record contractor performance job by job — quoted against actual, on time or not, rework, safety record. That picture can stay internal to safety or feed your supplier reviews. This is usually the easiest first move, because the boundary is uncontested.
The most expensive time on a paper site is a full crew waiting for one signature.
Digital routing turns a walk-and-wait into a notification, across every approval a crew actually stands still for:
- Permit-to-work approval
- Permit extensions
- Pre-task risk assessment approval
- Equipment inspection sign-off
- Contractor clearance to start
- Stop-notice close-out
The same choice exists for competencies, asset condition history and safe-work procedures — the Integration Pack maps all of them for your specific system.
The integration, in plain terms
No duplicate data capture. No second HR system to maintain. Your ERP stays authoritative; we subscribe to it.
| Data | Source layer | What it enables |
|---|---|---|
| Employees, org units, job titles, supervisors | ERP HR / payroll | Risk assessments, training matrices and permit authority stay aligned to your current org structure |
| Cost centres, sites, shafts, business units | ERP finance | So incident cost and lost-time exposure report against your real cost hierarchy |
| Equipment and asset registers | EAM / ERP | Pre-use inspections, isolation and lockout, and equipment-linked incidents |
| Contractors and vendors | ERP procurement | Contractor compliance gating — no valid documentation, no site access |
| Training records and qualifications | ERP HR / LMS | Competency currency checks before high-risk work is authorised |
| Shift patterns and rosters | ERP HR / time & attendance | Fatigue exposure and correct incident time attribution |
| Production volumes | Technical / operational layer | Normalised safety rates — incidents per million tonnes, not raw counts |
| Data | Destination | What it unlocks |
|---|---|---|
| Incident cost and downtime | ERP finance | True cost of safety failure, posted against the cost centre that carried it |
| Corrective action spend and procurement requests | ERP procurement | Remedial work becomes a tracked purchase, not an email |
| Equipment defects and unsafe conditions | EAM | A field observation becomes a maintenance work order automatically |
| Competency gaps and expiry warnings | ERP HR / LMS | Training scheduled before the certificate lapses, not after the audit |
| Compliance and regulatory status | ERP / BI layer | Board and DMRE reporting from live data instead of a month-end spreadsheet |
Data we consume
We consumeEmployees, org units, job titles, supervisors
Source layer: ERP HR / payroll
What it enables: Risk assessments, training matrices and permit authority stay aligned to your current org structure
Cost centres, sites, shafts, business units
Source layer: ERP finance
What it enables: So incident cost and lost-time exposure report against your real cost hierarchy
Equipment and asset registers
Source layer: EAM / ERP
What it enables: Pre-use inspections, isolation and lockout, and equipment-linked incidents
Contractors and vendors
Source layer: ERP procurement
What it enables: Contractor compliance gating — no valid documentation, no site access
Training records and qualifications
Source layer: ERP HR / LMS
What it enables: Competency currency checks before high-risk work is authorised
Shift patterns and rosters
Source layer: ERP HR / time & attendance
What it enables: Fatigue exposure and correct incident time attribution
Production volumes
Source layer: Technical / operational layer
What it enables: Normalised safety rates — incidents per million tonnes, not raw counts
Data we return
We returnIncident cost and downtime
Destination: ERP finance
What it unlocks: True cost of safety failure, posted against the cost centre that carried it
Corrective action spend and procurement requests
Destination: ERP procurement
What it unlocks: Remedial work becomes a tracked purchase, not an email
Equipment defects and unsafe conditions
Destination: EAM
What it unlocks: A field observation becomes a maintenance work order automatically
Competency gaps and expiry warnings
Destination: ERP HR / LMS
What it unlocks: Training scheduled before the certificate lapses, not after the audit
Compliance and regulatory status
Destination: ERP / BI layer
What it unlocks: Board and DMRE reporting from live data instead of a month-end spreadsheet
No ERP? Nothing here is a prerequisite.
RiskSafetyAI holds its own people, sites, contractor and equipment records, and runs complete without a single integration. If you have an ERP, we read from it so you maintain that data once. If you don’t — or if yours doesn’t hold what we need — you maintain it here, and connect later when it suits you.
Integration methods
How we build your integration. Most implementations use two or three of the following — chosen in the integration review against your estate, and built during phase one.
REST API
Documented, versioned API for read and write against any modern ERP or middleware layer. The default for cloud-hosted SAP, Dynamics, IFS and NetSuite.
Identity federation
SSO via SAML 2.0 and OpenID Connect; automated user provisioning and deprovisioning via SCIM. Your Entra ID or Okta directory stays the single gate for access — when someone leaves, they lose site access the same day.
Scheduled data sync
Secure file-based or database-level synchronisation for older on-premise estates where an API is not available or IT policy forbids one. Nightly, hourly or on-demand.
Middleware and iPaaS
Where you already run an integration layer — SAP Integration Suite, Azure Logic Apps, MuleSoft, Boomi — we connect through it rather than around it.
Offline-first field capture
The layer nobody else in your stack has. Full functionality underground and in remote pits with no connectivity; queued sync on reconnection. English, isiZulu, French, German and Arabic ship as standard, and we configure any other language your crews need.
This is not a system replacement project
Nobody replaces an ERP by accident. It is an eighteen-month, eight-figure programme with a steering committee, a change management workstream and a realistic chance of failure. That is the fear behind the question “are you an ERP?” — and it is a reasonable fear.
RiskSafetyAI is the opposite kind of project. It deploys in weeks, not quarters. It runs as operating expenditure, not capital. It requires no change to your chart of accounts, your master data model or your finance processes. And its effect on your ERP investment is entirely additive: the compliance obligations your ERP cannot discharge get discharged, and the safety data your ERP cannot capture starts arriving in it.
| Dimension | ERP replacement | RiskSafetyAI |
|---|---|---|
| Timeline | 12–24 months | 4–8 weeks to first site live |
| Cost structure | Capital programme | Operating subscription |
| Business disruption | Finance, supply chain, HR, payroll | SHEQ team and field users only |
| Master data | Rebuilt and migrated | Read from your existing ERP |
| Risk if it goes wrong | The business stops | A module is switched off |
| Effect on current ERP | Decommissioned | Extended and better utilised |
Timeline
12–24 months
4–8 weeks to first site live
Cost structure
Capital programme
Operating subscription
Business disruption
Finance, supply chain, HR, payroll
SHEQ team and field users only
Master data
Rebuilt and migrated
Read from your existing ERP
Risk if it goes wrong
The business stops
A module is switched off
Effect on current ERP
Decommissioned
Extended and better utilised
The full detail lives in the Integration Pack
Field-by-field data flows, the complete ERP comparison, the architecture document, and a system-specific integration overview — IFS, SAP, Dynamics, Maximo — sent to your IT lead before we meet.
Request the Integration PackSystems we integrate with
The integration pattern is designed for the systems below — and your connector is built against your actual estate during phase one, read-only first. If your system isn’t listed, the integration review will tell you honestly whether it’s a straightforward build, a sync, or a problem. And if you don’t run one at all, that’s not a problem either.
Most common in mining
- SAP
- IFS
- SYSPRO
- Dynamics 365
- Sage
- Oracle
- Maximo
- Ellipse
- Pragma On Key
Enterprise ERP · 6
- Dynamics 365Microsoft · Finance & Operations
- IFSCloud
- InforCloudSuite · LN
- OracleE-Business Suite
- OracleFusion Cloud ERP
- SAPS/4HANA · ECC
Mid-market ERP · 10
- Acumatica
- Dynamics 365Microsoft · Business Central
- Embrace ERP
- EpicorKinetic
- NetSuiteOracle
- Odoo
- Pronto Xi
- SageX3 · 300 · Intacct
- SAPBusiness One
- SYSPRO
Asset management · 5
- EllipseHitachi
- Infor EAM
- MaximoIBM
- Pragma On Key
- SAP PM/EAM
Identity and productivity · 4
- Entra IDMicrosoft
- Google Workspace
- Microsoft 365
- Okta
Operational data
Fleet management and equipment monitoring platforms, LIMS and process historians via API or scheduled sync.
Product names are the trademarks of their respective owners and are listed here to describe compatibility only.
Questions we get asked in every first meeting
Most major ERPs ship one — SAP EHS, IFS Health, Safety & Environment, Infor EAM’s HSE module, Dynamics through a partner add-on. SAP EHS is a strong compliance and reporting module, and if it is already configured, populated and adopted by your field teams, you may not need us. In practice the gap we are called into is execution at the face: offline capture underground, a supervisor completing an inspection on a phone with no signal, a permit to work issued and closed in the field, an isiZulu-language risk assessment a rock drill operator can actually read. SAP EHS was built for the desk. We were built for the shaft, and we feed the desk.
No. Your ERP remains authoritative. We sync employees, org structure and roles from it. Nobody is captured twice, and when a change is made in the ERP it propagates to us.
Then you re-point the integration. Because we never became your system of record for people, finance or assets, an ERP migration does not put your safety history at risk — which is more than can be said for the safety data currently sitting in ERP custom fields and Excel.
Bring them to the integration review. The conversation they usually expect is a data-sovereignty argument; the conversation they get is SSO, SCIM, data residency, encryption and audit logging. Most IT objections we meet dissolve in that first hour.
No. Those are the technical and operational layer. We consume production volumes from them so safety statistics can be normalised against tonnes, and we consume equipment data so incidents attach to real assets. We have no ambition to plan your mine.
Then you have made a considered choice in our layer, and the honest answer depends on what you use it for. Those platforms reach well beyond us — ESG and carbon reporting, governance and enterprise risk, occupational health, product stewardship — and we have no ambition to replace any of it. But we are not a smaller version of them either. They grew out of enterprise compliance reporting; we grew out of the working place. Every form is captured offline underground and assessed by AI on submission rather than filed, in whichever languages your crews actually speak — a different product for a different problem, not a cheaper seat at the same table. Where we overlap is execution in the field: the risk assessment completed with no signal, the permit closed at the face, the assessment a rock drill operator can actually read. If that is the part that is not landing, we sit alongside what you already have and feed it. If you are evaluating the whole layer fresh, we are a genuine competitor — and we would expect to differ on offline capability, deployment time, African language support and total cost.
Thirty minutes will settle it
Bring your IT lead and your SHEQ manager. We will map your specific stack — ERP, EAM, and whatever else is in the estate — and tell you exactly what integrates, what needs a sync, and what doesn’t connect. If the answer is that you don’t need us, you’ll hear that too.
Book a demo