Secure AI Agent · Occupational Health and Safety
AI applied to your safety program
A specialist configured to keep the hazard inventory, the training, PPE, and medical surveillance matrix with due dates in plain view, and the inspection and committee calendar — between the safety professional's visits.
Companies that trust UPX
Capabilities
What the Safety Agent can do
Main areas of work for the Occupational Health and Safety AI Agent in your operation.
Hazard inventory
Keeps identified hazards by area and role, with the date of the last assessment.
Written program skeleton
Organizes the structure and data that feed the document, for the safety professional to author.
Training matrix
Tracks who completed which training, when it expired, and who is overdue.
PPE and exam tracking
Records issuance against a signed receipt and tracks each medical surveillance due date.
Incident investigation
Builds the case with a timeline and evidence and flags the applicable legal deadline.
Inspection calendar
Keeps inspections, safety committee meetings, and drills with findings chased to closure.
Skills
Capabilities that compose the specialist
Skills add specific capabilities to the Secure AI Agent according to the processes it needs to execute.
- Due-date ladder
- Cross-references training, PPE, and exams per worker and alerts ahead of the date.
- Incident workup
- Builds the timeline with evidence and flags the clock that is running.
- Inspection findings
- Tracks each finding to closure, with an owner and a date.
From inventory to inspection, with the safety professional in charge
Connect your operation's spreadsheets and calendar, define what the agent can do, and let the tracking happen within your policy's limits.
- Connect the operation's sources
- The agent works with what the company already records.Connect the worker roster, the training and PPE log, email, and the calendar so the Secure AI Agent has access only to what it needs to track deadlines and findings.
- Ask for the work
- Talk to the agent in natural language.Request who is overdue on training, the workup for an incident, or what an inspector would see first. The agent understands the context, applies the configured skills, and follows the operation's rules.
- The agent organizes. The professional signs.
- Technical documents are not the agent's.The agent keeps the inventory, flags due dates, and works up the case within the defined permissions. Authoring and signing the written program, medical protocol, and reports stay with the licensed professional.
01
02
03
Integrations
Connected to the systems where your operation already records
The Safety Agent can query records and run tasks in the systems your team already uses.
Flow
What goes in, what the agent does, and what comes out
From raw operational records to a prioritized list of open items, following your operation's rules and permissions.
Inputs
- Workers and rolesCSV
- Training and examsAPI
- Incident reportDOC
Processing
Secure AI Agent
Processing the task
- Read
- Cross-reference by role
- Check the clockrunning
- Prioritize
Output
Completed
Open items prepared
- Overdue per worker
- Incident with the deadline flagged
- Open inspection findings
Control
You define how far the Agent can act
Different actions can operate with different autonomy levels, always within your operation's rules.
- 1
Query
Reads records and deadlines and answers with the reference.
- 2
Prepare
Builds the open-items list and works up the incident.
- 3
Request review
Waits for validation from the safety professional.
- 4
Execute
Performs the action within the defined limits.
Levels are configured per type of action, according to each operation's policy. Authoring and signing technical documents, filing an occupational accident report, and removing a worker from duty always stay with the licensed professional and the company.
Get started
Put a Secure AI Agent to work.
Start on the platform or choose the plan that fits the pace of your operation.
Security that can be verified.
Certifications and attestations
UPX maintains SOC 2 Type II and ISO 27001, with independent audit over its information security controls.
Privacy and regulation
- LGPD
- Operations follow Brazil's Law 13.709/2018. In AI Agent contracts, UPX acts as data processor; the legal basis remains with your company.
- Zero Data Retention
- A product policy, not a certification: all market AI models used by the Agents have ZDR under contract, and processed content is not retained after execution.
- Retention and deletion
- The retention policy is defined by contract. Once the contract ends, data is deleted within the agreed period.
Frequently asked questions
Common questions about the Safety Agent
What companies usually ask before putting an agent to support workplace safety.
Does the agent author the written safety program?
No. It keeps the hazard inventory, organizes the skeleton and the data that feed the document, and tracks deadlines. Authoring and signing are acts of a licensed professional — the safety engineer and the occupational physician.Does it replace the safety professional or consultancy?
No. This profile exists to keep the program alive BETWEEN visits: when the professional returns, they find an updated inventory, mapped due dates, and recorded inspection findings instead of rebuilding everything.Does it decide whether an accident is recordable?
It does not decide. It works up the case — timeline, evidence, likely classification — and flags the legal deadline that is running. Filing is the company's responsibility, backed by the occupational physician.How does due-date tracking work?
The agent cross-references training, PPE issuance, and exams per worker and alerts ahead of the date, not after. Blocking or removing someone who is overdue is the operation's decision — the agent makes the overdue visible.What does it say about an inspection?
It points out what is expired, lacks evidence, or has no assigned owner — that is, what gets exposed first in an inspection. It is a reading of your own records, not a prediction of what the inspector will do.Is worker health data used to train models?
No. Content processed by Secure AI Agents is not used to train UPX models or third-party models. Health data is sensitive and receives the same restricted handling as the rest of the operation.How long does it take to go live?
It depends on what is already recorded. The starting point is loading the worker roster and the training log, then expanding the scope as the hazard inventory is validated by the safety professional.
Custom agents
We build and manage the agent for your safety program
From the hazard inventory to the inspection calendar: we build the agent that keeps the program alive between visits.