Heartpace AI

Three agents. One Heartpace.

A transparent, role-based AI layer for HR work: each agent reads approved Heartpace data, drafts the work, asks for approval, and logs every action.

Heartpace AI

What stays human

Outcome

HR Admin Agent

for HR

EU Pay Transparency Directive reporting · Equal Pay analysis with job-evaluation context

Manager Coach Agent

for Managers

1:1 and quarterly Talks prep packs · OKR drafting and progress nudges

Employee Service Agent

for Employees

Vacation, leave, sick day balance and requests · Policy lookup in any language

Reads

Drafts

Human approval

Outcome

Plain-language AI

What a Heartpace agent actually does.

It is not a black-box chatbot. A Heartpace agent is a role-specific assistant with scoped tools, permission-aware context, human approval gates and audit logs.

It reads approved Heartpace data

Employee records, policies, documents, salary data, OKRs, surveys and workflows — only within the user's permissions.

It drafts work, not final decisions

Reports, talking points, onboarding tasks, salary review notes and policy answers are prepared for humans to review.

It asks for approval

Sensitive actions like pay changes, reports, employee messages and escalations stay behind explicit approval gates.

It leaves an audit trail

Every tool call, source, draft, approval and workflow step is logged for GDPR, ISO 27001 and internal review.

Important for HR buyers

Agents do not replace HR judgement. They reduce admin, prepare better decisions and make sure the work is traceable.

Human in control
HR Admin Agent · for HR

Equal pay, GDPR and audit-ready work in minutes.

A governed assistant for HR compliance work. It runs Equal Pay analyses, drafts EU Pay Transparency Directive reports, redacts GDPR access requests and triages whistleblowing intake — every action audit-logged and reversible.

ScenarioAnna runs Q1 Equal Pay analysis

Anna · HR Manager · 487 employees

What it does

  • EU Pay Transparency Directive reporting
  • Equal Pay analysis with job-evaluation context
  • GDPR data access request handling
  • Whistleblowing intake triage and redaction
  • Policy gap analysis vs Nordic labor law
  • DORA incident drill orchestration
  • Audit trail and signature workflows

What stays human

Pay adjustments, terminations, signed reports, anything regulator-bound — always require human approval.

See the role page →

HR Admin Agent · for HR

Real cases it can solve

Real case

EU Pay Transparency report before the deadline

approval-gated
Trigger

HR asks for Q1 pay transparency readiness across 100+ employees.

Reads

salary mapping, job architecture, role groups, market salary data

Drafts

A report with pay gaps, explanations, missing data and suggested remediation notes.

Human approval

HR reviews, CFO approves, union/committee pack is generated only after sign-off.

Outcome

Directive-ready evidence pack with audit trail, source links and next actions.

pay.equal_gap.scanpay.transparency.draftaudit.approval.request

Real case

Whistleblowing intake without exposing sensitive data

approval-gated
Trigger

A new anonymous case is submitted and HR needs a safe first triage.

Reads

case category, policy, timeline, attachments metadata

Drafts

A redacted summary, severity classification and recommended handler routing.

Human approval

Only the assigned HR/legal owner can open the case details and take action.

Outcome

Faster triage while protecting anonymity and keeping legal documentation intact.

whistle.case.triagegdpr.redactaudit.log.write
Manager Coach Agent · for Managers

Better Talks, faster. Grounded in your data.

Drafts 1:1 prep cards with goals, 360 context and privacy-safe anonymous survey trends. Recommends salary adjustments grounded in Market Salary Data. Detects burnout and disengagement risk before they become exits.

ScenarioErik prepares for a 1:1

Erik · Engineering Manager · 8 reports

What it does

  • 1:1 and quarterly Talks prep packs
  • OKR drafting and progress nudges
  • 360 evaluation summarization
  • Salary review recommendations with market data
  • Anonymous survey trend detection with privacy thresholds
  • Burnout and disengagement early warnings
  • Coaching tips tailored to manager experience level

What stays human

Salary changes, performance ratings, written feedback to employees — drafted by AI, approved by humans.

See the role page →

Manager Coach Agent · for Managers

Real cases it can solve

Real case

Manager Talk prep that actually uses the data

approval-gated
Trigger

A manager opens tomorrow's 1:1 and asks what to focus on.

Reads

previous Talks, OKR progress, anonymous team survey trends, 360 feedback

Drafts

A structured agenda with questions, risks, praise points and follow-up suggestions.

Human approval

Manager edits the agenda and decides what is shared with the employee.

Outcome

Better prepared conversations and more consistent leadership quality.

performance.talks.prepareokr.status.readsurvey.signals.summarize

Real case

Salary review guidance for fair manager recommendations

approval-gated
Trigger

Salary review opens and managers need consistent recommendation support.

Reads

salary band, market salary data, performance notes, budget rules

Drafts

A recommendation range with rationale, fairness warnings and manager talking points.

Human approval

Manager adjusts, HR calibrates, final salary decisions remain human-approved.

Outcome

Less bias, less HR bottleneck and faster salary review cycles.

salary.review.contextmarket.salary.compareapproval.route
Employee Service Agent · for Employees

Instant answers across the employee lifecycle.

An always-on service desk for every employee, in their language. Answers vacation balance, policy, and benefits questions. Starts workflows for onboarding, eSign, OKR drafts, and Talks prep. Routes anything sensitive to the right human.

ScenarioMaja asks a quick question

Maja · Senior Engineer · Stockholm

What it does

  • Vacation, leave, sick day balance and requests
  • Policy lookup in any language
  • eSign and document retrieval
  • Onboarding tasks and check-ins
  • OKR draft suggestions
  • 1:1 prep notes for the employee side
  • Hand-off to HR for sensitive topics

What stays human

Anything outside published policy is escalated to HR. Sensitive topics (harassment, whistleblowing) go directly to a human.

See the role page →

Employee Service Agent · for Employees

Real cases it can solve

Real case

Vacation and policy answers without HR tickets

approval-gated
Trigger

An employee asks how many vacation days remain and what can be carried over.

Reads

vacation balance, contract, local policy, manager hierarchy

Drafts

A plain-language answer with policy source and an optional vacation request.

Human approval

Manager approves the leave request in the normal Heartpace workflow.

Outcome

Employee gets an instant answer; HR avoids another routine ticket.

vacation.balance.readpolicy.answerworkflow.leave.request

Real case

Onboarding support from day one to day ninety

approval-gated
Trigger

A new hire asks what to complete before their first day.

Reads

onboarding template, documents, manager tasks, role profile

Drafts

A personal checklist, document reminders and questions for the first manager check-in.

Human approval

HR owns the template; manager confirms role-specific tasks.

Outcome

New hires get clarity without waiting for HR or their manager to answer everything.

onboarding.plan.readdocuments.requestmanager.checkin.schedule

What the agents cannot do

Heartpace AI agents are governed assistants, not autonomous decision-makers. Here is what they will never do without explicit human approval.

  • Make salary changes or process payments
  • Expose individual eNPS or anonymous survey responses
  • Send external communications on behalf of HR
  • Delete or alter employee records
  • Override role-based access controls

See the three agents in a real workflow.

Bring your HR, manager and employee scenarios. We'll show how the agents support them with approval and auditability.