Skip to content

Intelligent HSE

AI that understands your operations — and respects your controls.

Capture → Understand → Check → Act. Green Codes Pro turns photos, messages and records into actionable safety intelligence, while qualified people stay in charge of every decision.

Book an AI demo

How it works

From a workplace observation to actionable intelligence.

  1. 1

    Capture

    A photo, voice note or WhatsApp message arrives from site, or a question is typed in Arabic or English.

  2. 2

    Understand

    The assistant describes potential hazards, suggests the record type and reads the related history you may see.

  3. 3

    Check

    Findings can be matched with requirements from your own regulation library and procedures.

  4. 4

    Act

    Draft observations, actions or reports are prepared. Nothing is saved until a person reviews and confirms.

Try it

AI photo safety review

A site photo becomes a structured, reviewable list of potential hazards and suggested actions.

AI photo safety review

Illustrative demonstration with sample data.

Step 1 of 5: Capture. A photo is taken on site or uploaded from a phone.

123
Site photo · Tower B, Level 4

Potential hazards 0

Responsible by design

Built-in guardrails, not afterthoughts.

AI in Green Codes Pro supports professional judgement. It does not replace qualified review.

Permission-aware

Answers are built only from records the signed-in person could already open; confidential lists return totals only.

Draft-only suggestions

Suggested observations, actions and findings stay drafts until a person confirms them.

Data-category routing

Personal, medical, commercial or legal data goes only to AI connections approved for it.

Your choice of provider

Connect supported providers where configured, with an AI governance page to manage models, budgets and usage.

Works without AI

If no provider is configured or one is unavailable, pages still answer from the platform’s own figures and explain why.

Cited answers

Answers over your documents and help articles cite the passages they used.

AI across the platform

Where AI assists in Green Codes Pro

AI assistance is built into the modules where it helps people work faster. Every suggestion is reviewed by your team before anything is recorded.

Modules with AI assistance: 30

AI suggestions support qualified people — they never replace professional review, and every decision stays with your team.

Safety operations

7

Safety Observations

Where an AI provider is configured, AI can read a site photo and suggest a title, type and risk level for the reporter to review; only empty fields are filled and nothing is saved until the person applies it. AI can also summarise an observation and look for similar past observations.

Open module: Safety Observations

AI features 2

  • AI suggestions from a photoWhere AI is configured, the reporter can ask for a suggested title, type and risk from a site photo; suggestions fill only empty fields and are reviewed before applying.
  • AI summary and similar observationsAI can summarise an observation and check whether similar observations have been reported before, to help spot repeat issues.

Incident & Near-Miss Management

Where AI is configured, it can suggest an incident level from the recorded impact, summarise an incident and find similar past incidents; incident search can match by meaning when an embedding model is assigned. People review and decide on every suggestion.

Open module: Incident & Near-Miss Management

AI features 2

  • Search incidents by meaningDescribe what happened in your own words to find similar incidents; matching by meaning works where an embedding model is assigned, otherwise it searches by words.
  • AI incident suggestionsWhere AI is configured, it can suggest a level, summarise the incident and list similar incidents; each suggestion is accepted or dismissed by a person.

Incident Investigation & Root Cause Analysis

Where AI is configured, an investigation assistant lists what is recorded and what is missing and offers possible contributing factors and questions, clearly labelled as hypotheses. At completion AI can suggest preventive actions that an investigator raises as draft actions or dismisses with a reason, and can draft a learning pack for review. The official root cause remains a person's decision.

Open module: Incident Investigation & Root Cause Analysis

AI features 3

  • AI investigation assistantWhere AI is configured, the assistant lists gaps in what is recorded and offers possible contributing factors and questions to ask, labelled as hypotheses.
  • AI preventive action suggestionsWhen switched on by the organization, AI proposes preventive actions from the recorded facts and causes; an investigator raises each as a draft action or dismisses it with a reason.
  • AI learning pack draftWhere AI is configured, draft lessons learned, a safety alert, a toolbox talk and a short quiz from an investigation for a person to review; nothing is published automatically.

Electronic Permit to Work

Where AI is configured, the assistant's permit review explains the permit's own readiness verdict, gas-test and isolation status, and suggests controls and JSA points for people to consider. The permit lifecycle stays rule-based and AI changes nothing on the permit.

Open module: Electronic Permit to Work

AI features 1

  • AI permit reviewWhere AI is configured, the assistant explains a permit's readiness, gas-test and isolation status and suggests controls for review, without changing the permit.

Monthly HSE Reports

Where AI is configured, AI acts as a second reader of the submitted figures alongside rule-based checks, flagging items for reviewers; it never proposes numbers.

Open module: Monthly HSE Reports

AI features 1

  • Rule and AI reviewRule-based checks flag unusual figures for reviewers; where AI is configured, it adds a second reading but never proposes numbers.

Safety Alerts

Where AI is configured, "Draft with AI" proposes the alert text for the author to edit; a template is used when no model is available.

Open module: Safety Alerts

AI features 1

  • Draft with AIWhere AI is configured, propose the alert text for the author to edit, with a template fallback.

Risk & compliance

3

Risk Assessment & Risk Register

Where AI is configured, the risk assistant suggests controls from the organization's own library for a hazard, similar risks and possibly applicable requirements. It never produces a rating: ratings come only from the published matrix.

Open module: Risk Assessment & Risk Register

AI features 1

  • AI control suggestionsWhere AI is configured, get suggested controls from your own library, similar risks and possibly applicable requirements — never a rating.

Management of Change

Where AI is configured, five AI tools help find and summarise changes, find similar past changes, suggest impact areas and surface lessons; none of them approves, rates or closes a change.

Open module: Management of Change

AI features 1

  • AI change assistanceWhere AI is configured, find and summarise changes, see similar past changes, get suggested impact areas and lessons — AI never approves, rates or closes.
Pilot

Legislation & Compliance

AI proposes a regulation's clause structure for human review, classifies what changed between versions, evaluates records against requirements written in words and checks site photos against photo rules — every result is a potential finding that an authorised person confirms. AI never states that a record is compliant.

Open module: Legislation & Compliance

AI features 4

  • AI-proposed clause structure with human reviewAI proposes parts, chapters, articles and clauses; reviewers approve, correct text and repair layout before anything is used.
  • Compliance check on recordsA compliance panel on observation, incident, corrective action and NCR pages checks the record on demand against applicable requirements.
  • Photo compliance checkCheck a site photo against the photo rules defined for your standards and review the potential findings before raising anything.

People & competency

5
Pilot

Workforce Intelligence

AI assists by explaining why a worker's operational risk score is what it is, from the same factors shown on the page. It is read-only and decides nothing.

Open module: Workforce Intelligence

AI features 1

  • AI explanation of worker riskAsk the assistant why a worker is at a given risk level; it reads the same factors and changes nothing.
Pilot

Safety ePassport

AI assists by explaining why a passport shows its state, what expires next and how to fix it, in operational wording only. It is read-only.

Open module: Safety ePassport

AI features 1

  • AI passport explanationAsk why a passport is restricted or suspended; the assistant answers with reasons, next expiries and remediation, and changes nothing.

Training & Competency

AI assists with training needs, competency gaps, expiring training, course suggestions and explanations of an eligibility answer, using facts the matrix and eligibility engine already computed. People assign, assess and decide.

Open module: Training & Competency

AI features 1

  • AI training assistanceAI suggests training needs, competency gaps, expiring training and courses, and explains an eligibility answer from computed facts; people decide.

Learning Management & Examinations

AI assists course authors with an outline and lesson drafts and drafts exam questions as drafts only; a person reviews, edits and publishes. During a protected exam the assistant answers nothing about exam content.

Open module: Learning Management & Examinations

AI features 1

  • AI-assisted course and question draftingAI suggests a course outline, lesson drafts and exam questions as drafts that a person edits and publishes; nothing AI writes reaches a learner on its own.

Occupational Health & Hygiene

AI assists authorised users with summaries of examinations and cases, suggested questions, operational explanations of fitness, return-to-work note drafts and exposure trends, routed only to AI providers approved for this category of data where configured. AI never makes a fitness decision, diagnosis or reportability call; clinicians decide.

Open module: Occupational Health & Hygiene

AI features 1

  • AI support for cliniciansAI drafts summaries, questions and return-to-work notes and explains exposure trends for authorised users; it never decides fitness or diagnosis.

Operations & assets

4

Equipment & Asset Management

AI assists by explaining an asset's facts and readiness verdict, answering fleet questions from the register the reader may see, and offering troubleshooting help from the asset's history. The readiness rules decide; AI never sets the verdict.

Open module: Equipment & Asset Management

AI features 1

  • AI equipment assistantAI explains an asset's readiness, answers fleet questions and suggests troubleshooting steps from its history; it never sets readiness.

Contractor Management

AI assists with a contractor summary, a list of expiring documents and an explanation of how an evaluation score was reached. It is advisory; people approve, score and sanction.

Open module: Contractor Management

AI features 1

  • AI contractor assistantAI summarises a contractor's standing, lists expiring documents and explains how an evaluation score was reached; people decide.

Quality Management

AI assists with NCR facts and similar NCRs, root-cause suggestions, inspection finding summaries, defect classification and quality trends, read through the same visibility as the registers. It is advisory; people decide.

Open module: Quality Management

AI features 1

  • AI quality assistantAI suggests root causes, summarises inspection findings, classifies defects, finds similar NCRs and describes trends; people decide.

WhatsApp & Mobile Field Reporting

Where configured, AI describes photos sent by WhatsApp, transcribes voice notes and reads them back, and suggests the kind of report; the sender confirms before anything is filed.

Open module: WhatsApp & Mobile Field Reporting

AI features 2

  • Photos read and attachedPhotos are attached to the record as evidence and, where AI is configured, described before the sender confirms — a clean site is recorded as a valid positive observation.
  • Voice notes transcribedVoice notes are transcribed where a speech provider is configured and read straight back to the sender for confirmation.

Security

1

Security Operations

AI assists the control room with a situation summary, ticket summaries, triage suggestions and operational pattern signals drawn only from records the reader may see. It cannot dispatch, close, suspend, change access, start a muster or judge a person; operators decide.

Open module: Security Operations

AI features 1

  • AI copilot and insight signalsAI assists with the control room situation, ticket summaries, gate-decision explanations and operational patterns such as repeated gate faults; it is read-only and people take every decision.

Facilities

1
Pilot

Total Facilities Management

AI assists the helpdesk with facility facts, a job finder and request triage suggestions (category, priority, possible duplicates and, where AI is configured, a reading of the description and photos). It is read-only; the helpdesk decides.

Open module: Total Facilities Management

AI features 1

  • AI helpdesk assistanceAI suggests a category, priority and possible duplicates for a request and answers facility and job questions from records the reader may see; people decide.

Environment & ESG

1
Pilot

ESG & Sustainability

AI assists with ESG questions, drafts narrative and disclosure text, explains variances and reads figures from bills or documents; people review drafts and save them, and AI wording never claims an SDG has been achieved.

Open module: ESG & Sustainability

AI features 1

  • ESG AI assistanceAI answers ESG questions, drafts disclosure text and narrative, explains variances and target slippage, and reads figures from documents — as drafts a person saves.

Intelligence & analytics

4

Risk Intelligence & Smart Alert Center

Where AI is configured, AI writes executive briefings, explains a signal and suggests next steps, and breaks ties in ordering within a priority — never changing the priority. Facts are computed first, numbers and references not in the evidence are removed, and a suggestion becomes a corrective action only through a draft a person confirms. Every AI feature has a rule-written fallback; early warnings are trend-based, not predictive.

Open module: Risk Intelligence & Smart Alert Center

AI features 3

  • AI executive briefingWhere AI is configured, a briefing for the organization, a project or a contractor, written from computed facts with unsupported numbers removed.
  • Explain and suggestOn request, AI explains a signal and suggests next steps; a suggestion becomes a corrective action only after a person reviews and confirms the draft.
  • Ask about your risk dataAsk questions on the dashboard; answers come from the reader's own signals and records, cite their sources and say when they cannot answer.

AI Assistant & Intelligent HSE

The assistant, AI Reports, AI Insights, AI Vision, AI Compliance, AI Knowledge and AI Automations assist people; findings are worded as potential and every record write goes through a draft that a person confirms. External AI providers are used only where the client has configured them.

Open module: AI Assistant & Intelligent HSE

AI features 20

  • AI Assistant on every pageA side panel available on every page that knows which page and record you are looking at, with private conversation history per user.
  • Ask anything about your recordsAsk in plain English or Arabic for lists, counts and breakdowns of records across modules — such as open permits or overdue actions — and get answers with verified record references.
  • Opens pages for youSay “open the permits page” or “take me to” a record reference and the assistant opens it with the right filters — only pages you may open, never an action such as approve or delete.

Report Center

The assistant can list the datasets you may use, find and run saved reports, and draft a table, chart or dashboard that opens in the designer for you to check and save; it is never told about confidential or restricted fields.

Open module: Report Center

AI features 2

  • AI report draftingAsk the assistant to draft a table, chart, dashboard or a change to an existing report; it opens in the designer and nothing is saved, shared or sent for you.
  • AI insights on a reportAI explains a report's results and highlights what stands out, from the figures the platform calculated.

Platform & governance

4

Help Center, Support & Video Academy

A help-center assistant answers only from published articles with verified citations and says honestly when a topic is not covered; a ticket summariser helps agents using only what they may see.

Open module: Help Center, Support & Video Academy

AI features 2

  • Help-center AI answersAnswers come only from published articles with verified citations; when nothing is found the assistant says so and offers a ticket.
  • Ticket summaries for agentsAI summarises a ticket from what the reader may see and writes nothing back.

See the AI assistant on your own scenarios.

AI features depend on the provider configuration agreed for your deployment.

Book a demoRequest a quotation