Sense AI — also known as AMI (pronounced “Ah-Mee”, as in Miami) — is the intelligent assistance layer within the Sense Workplace ecosystem.
AMI is designed to help users analyse data, automate tasks, and generate useful outputs across multiple products, including HR, ATS, and Presence.
It works alongside users to improve efficiency, reduce manual effort, and support better decision-making — while keeping users in control of outcomes.
What AMI can do
AMI agents support a wide range of tasks depending on the product and data they are connected to, including:
analysing workforce, attendance, and operational data
generating and refining document content
processing payroll documents such as payslips and P11Ds
reviewing expense claims against company policy
supporting recruitment workflows such as interview preparation and candidate evaluation
analysing task and activity data from operational systems
Each AMI capability is designed to support a specific workflow using the data available in that part of the system.
How AMI works
AMI operates as a combination of:
AMI workspace agents (specialised agents you can select in the AMI workspace)
Embedded agents (contextual AMI features embedded within products)
Each is designed for a specific type of task and uses only the data available to the user.
AMI workspace agents
AMI includes a dedicated workspace where users can select and interact with specialised agents.
Access via:
Product switcher > AMI
Each agent is designed for a specific purpose and uses data from a particular product area.
Examples include:
Doc Sender (HR)
Sends document templates to employees for signing or review using filters such as department or locationHR Calendar (HR)
Analyses attendance, work patterns, anomalies, and time offPayroll (HR)
Validates, processes, and uploads payroll documents such as payslips and P11DsQuizzy (ATS)
Generates pre-screening interview questions based on job requirementsInterviewer (ATS)
Conducts structured interviews, evaluates responses, and records resultsTechTasksTracker (Presence)
Provides analytics and insights on task activity, performance, and anomaliesGeneral Assistant
Answers general questions that do not require access to system-specific data
Each agent can clearly define:
what it can do
what data it can access
how it supports a specific workflow
🖊️ Note: Some AMI workspace agents also appear as shortcuts in the relevant product. For example:
HR Dashboard > Documents > AMI icon → Payroll agent
HR Dashboard > Calendar > AMI icon → HR Calendar agent
🖊️ Note:
Some AMI workspace agents (such as Doc Sender) are provided as library workflows and must be installed and deployed via Sense Automate before they are available.
See the documentation for each agent for specific setup requirements.
Embedded agents (contextual AMI features embedded in products)
AMI is also built directly into specific features across Sense Workplace products.
These features are accessed from within the relevant module and are designed to work automatically with the data and workflow in that area.
Some contextual AMI features may open the AMI workspace to perform the analysis, while others operate directly within the product interface.
Examples include:
Reports (HR) → Analyse with AMI
Document templates (HR) → Document drafting with AMI
To Do > Expense claims (HR) → policy-based validation during approval (Expenses with AMI)
Vacancies > Summarise (ATS) → summarises key information about a job vacancy
🖊️ Summary Note:
Contextual AMI features are not independently selectable agents. They are triggered from specific areas of the system and use the relevant data and context automatically.
How AMI is accessed
Depending on the task, you can access AMI in different ways:
From the AMI workspace (AMI workspace agents)
Product switcher > AMI → select an agent
From within a product module
For example:
• HR → Reports → Analyse with AMI
• HR → Settings → Documents → Document drafting with AMI
• HR → To Do → Expenses with AMI
• ATS → Vacancies → Summarise with AMI
• HR → Documents / Calendar → AMI agent shortcuts
Some shortcuts open the dedicated AMI workspace with the relevant agent auto-selected, while others provide embedded AI functionality within the feature.
How AMI works with data
AMI operates within Sense Workplace’s security and permission model.
This means:
AMI only accesses data the user is authorised to view
each agent uses data from its relevant product area (e.g. HR, ATS, Presence)
Agents do not directly share or transfer raw data between each other
outputs are generated from available system data or user-provided inputs (such as uploaded files)
Some agents (such as Payroll or Interviewer) rely on data provided during the interaction, while others analyse existing system data.
Working with agents in AMI
When using the AMI workspace, you can select one or more agents to work together within the same conversation.
Each selected agent contributes based on its role and the data it has access to. This allows you to combine different capabilities to complete more complex tasks.
Using multiple agents together
In some cases, selecting multiple agents can help you achieve better results.
For example:
Use HR Calendar Event Insights to identify employees with specific attendance patterns or anomalies
Then use Doc Sender to assign a document or follow-up action to those employees
By combining agents, you can move from insight → action within a single workflow.
💡 Tip:
If you are unsure which agent to use, you can select multiple agents and ask your question — each agent will respond based on its capabilities.
Understanding conversation context
Each AMI conversation keeps context from previous messages.
This means:
agents use earlier prompts and responses to inform their answers
follow-up questions can build on previous outputs
If you want to start fresh without previous context influencing the response, select:
+ New chat
This is useful when:
switching to a different task
testing a new prompt
avoiding earlier responses affecting results
Reusing successful prompts
When you achieve the result you need in a conversation, you can ask AMI to provide the full prompt or instructions used to generate that output, then copy and save it to your own files for future use.
While you can revisit previous conversations using the sidebar, saving prompts separately makes them easier to reuse, standardise, and share with other administrators or managers.
This can be useful for:
recurring or scheduled tasks
repeated analysis or reporting
standardised tasks
For example, you might ask:
“Show me the full prompt used to generate this result”, or
“Give me the full prompt used to generate this result in code”
You can then copy 
and save this prompt for reuse when needed.
💡 Tip:
Saving and reusing prompts can help ensure consistent results while reducing the time and token usage required for repetitive tasks.
Your responsibility when using AMI
AMI is designed to assist with tasks, not replace user responsibility.
Users remain responsible for:
reviewing generated content (especially legal or contractual wording)
confirming actions before processing data
ensuring outputs are accurate and appropriate
making final decisions based on AMI results
🖊️ Important:
AMI provides assistance and recommendations, but all final responsibility remains with the user.
Best practice
Use clear and specific instructions when interacting with AMI
Select the correct agent or feature for your task (click Save changes to apply)
Review all outputs before taking action
Ensure underlying data (e.g. templates, policies, reports) is accurate
Use AMI to support decision-making, not replace it
💡 Tip: If you’re unsure what an agent can do, what it has access to, or how to use it, you can ask it directly.
Most agents can explain their capabilities, data access, and guide you through the task step by step.
Understanding AMI usage (tokens)
AMI usage is measured in tokens, which represent the amount of text processed during interactions.
Input tokens — the text you enter
Output tokens — the response generated by AMI
Your total usage is the combined total of input and output tokens
As a guide:
1 token is roughly 4 characters or ¾ of a word in English
You can monitor your usage using the token indicator in the AMI workspace, shown as a percentage with a progress bar (e.g. 
Used tokens 25%)
Each user has a monthly token allowance, which resets automatically each month.
If you reach your limit, AMI may prompt you to contact Sense Workplace Support to request additional usage.
🖊️ Note:
Token usage helps manage system performance and ensures consistent availability across all users.
Availability
AMI features are available depending on your organisation’s subscription plan.
General Assistant is available on all plans
Document template assistant (AMI in the editor) is available on Professional, Elite, and Enterprise plans
Advanced AMI workspace agents and workflow-enabled AMI features (such as Payroll agent, HR Calendar agent, Doc Sender agent, and Analyse with AMI) are available on Elite and Enterprise plans
🖊️ Note:
Access to specific AMI features and the data they use is also controlled by user roles and permissions. Administrators have full access, while other roles may have limited access depending on configuration.
Basic troubleshooting
If AMI is not behaving as expected, the following are common causes:
Issue | Possible cause |
AMI option not visible | Feature not included in your plan or restricted by access roles |
Incomplete or limited results | Missing or incomplete underlying data |
Unexpected or incorrect output | Prompt is unclear or lacks sufficient detail |
💡 Tip:
If you’re unsure how to phrase your request, try rewording your prompt with more detail, or ask the agent for guidance.
Summary
AMI is the intelligent assistance layer within Sense Workplace that helps automate and support tasks across multiple products, including HR, ATS, and Presence.
It works through:
AMI workspace agents
Embedded agents within product modules
By using AMI effectively, organisations can improve efficiency, reduce manual work, and gain faster insights — while maintaining full control and oversight.