One private assistant helps your people understand company policy, complete their training, find wellbeing support and speak up safely, while you gain answers that cite their source, documented training and a case workflow aligned with the EU Whistleblower Directive. Support, not surveillance.
Private — your manager and HR cannot read this conversation.
I'm worried about raising a concern about my manager. Is this really private, and what are my options?
Yes — this chat is private to you. Your manager and HR cannot read this conversation.
Based on your company policy, you can also report a concern confidentially, either:
Only you decide whether to make a report — I won't file anything for you. The policy says employees can report in good faith without fear of retaliation.
What's the main concern about your manager that's been weighing on you?
Hremia is an AI assistant, not a doctor or therapist. A human can always help.
In immediate danger? Call your local emergency number (112 in the EU) or a helpline.
Built for EU organisations in healthcare, logistics, retail, hospitality, manufacturing and the public sector.
Ask anything about the handbook, benefits or procedures. The AI answers from your company's own documents — with sources — in the employee's language.
How many remote days can I take?
Up to 2 remote days a week for your role.
From “Employee-Handbook.pdf”Turn your policies and regulations into courses. Employees learn with an AI coach that knows the material, take server-graded tests and get certified — HR sees who completed and who hasn't.
Employees ask the AI how many days of leave they have left, when their last holiday was or what their sick-leave rights are — answered from HR's own records, visible only to them. Fewer routine questions land on HR.
How many days of leave do I have left?
9 days left this year. Your last leave was 17–28 July.
From your HR record — visible only to youA calm, multilingual assistant for stress, workload and conflict — private from managers and HR, by database design.
Raise a concern named or fully anonymous, with a secure two-way channel and a fair, auditable response.
Every conversation — wellbeing or training — stays private. Managers and HR only ever see what an employee chooses to submit.
Built on the same rule as the rest of Hremia: support for people, never surveillance.
Upload a policy PDF and Hremia drafts multiple-choice questions, checking every answer against the passage it came from. HR edits, deletes or replaces any question before publishing, and employees answer in their own language.
Who approves overtime?
HR and department heads post news and events. Each person reads them in their own language, can confirm they have read them or reply to an invitation, and finds everything in their notifications even if they missed the alert. Nothing buzzes at night or at weekends unless it is urgent.
Summer party, Friday at 18:00
In the garden. Families welcome.
Short games in each person's language: giving change, finding a word, a quiz ladder that can include questions from courses already passed, mini sudoku, memory cards and guided breathing. The company sets the minutes per day. Nobody at the company sees who played or for how long.
8 of 10 minutes left today
Each employee picks a language once. The app, quizzes, announcements, games and emails follow it, in ten languages including Greek, Arabic, Bulgarian, Romanian and Turkish.
Employees ask in their own words, at any hour, in any of ten languages. Every answer comes from a document you uploaded or a record you keep — so the same question stops arriving at your desk.
Answered from your handbook, collective agreement or the law you upload — quoting the exact article.
Answered from the leave record your HR keeps, in a private conversation that HR and managers cannot read.
Answered from the employment record your HR files — stated as recorded, never calculated.
If the answer is not in your documents or your records, Hremia says so and points the employee to HR. It cannot cite an article that is not in its sources, and it never works out a figure the service has not recorded.
Wellbeing, psychological safety and trusted reporting are now operational, legal and reputational priorities.
of EU citizens reported an emotional or psychosocial problem in the past year
global productivity lost to depression and anxiety (WHO)
annual cost of poor mental health to employers (Deloitte, UK)
global employee engagement in 2025 (Gallup)
Talk in your own language about workload, stress, conflict or a policy question. The assistant listens — your manager and HR cannot read it. When you want action, you choose whether to file a named or anonymous report.
Learn moreReceive named and anonymous reports with a clear case trail, and see department-level wellbeing trends with strict anonymity thresholds — never individual chats, never identities below a safe group size.
Learn moreEmployees speak up in their own language; HR and DPOs handle named and anonymous reports in one secure inbox — with privacy-preserving insights, never raw chats.
Learn moreReports inbox
Named and anonymous reports for your organisation.
Customer service understaffing
Workload · Customer Service
Forklift near-misses during peak hours
Health & safety
Forklift near-misses during peak hours
HRM-NORTH001
Reporter
There have been several forklift near-misses during peak dispatch. People are rushing to hit targets and safety distance isn't kept. I'd rather report this anonymously.
HR
Thank you for raising this — safety comes first. We're reviewing peak-hour traffic routing with the warehouse leads and will update here.
The reporter reads this with their access code. HR answers without ever learning who they are.
From hospital wards to warehouse floors, Hremia meets people where they are — in their own language, on every shift, across borders and sites.
Designed for EU organisations of 100–2,000+ employees · 10 languages · multi-country, multi-site.
Earlier
Concerns surface sooner — before they turn into absence, conflict or resignation.
Fairer
Every report gets a consistent, auditable, human-reviewed response.
Provable
Compliance evidence for GDPR, the EU AI Act and ISO 45003 — by design.
Knowledge, training, wellbeing and trusted speak-up — consolidated, with compliance built in.
Multilingual chat, policy Q&A from your own documents, supportive reflection and next-step guidance.
Off by default. If an employee switches it on, the assistant keeps a few short notes between conversations — theirs to view, correct or delete at any time, and never visible to HR.
Report drafting, an anonymous two-way mailbox, routing, acknowledgement and deadlines.
Triage, replies, status and audit trail — a consistent, defensible record of fair handling.
Department-level wellbeing trends with strict anonymity thresholds. Never individual chats.
Upload your handbook and policies; the AI answers employees from the real documents — with citations — in their own language.
Build courses from your documents; assign mandatory or optional; employees learn by chatting with an AI tutor — completion, quizzes and certificates tracked.
Encryption, retention, audit logs and human oversight — aligned to GDPR and the EU AI Act.
Create your workspace, upload your handbook & policy documents, add departments, and invite employees, managers and DPOs. Set reporting routes and privacy notices.
Announce Hremia to your people, run a short manager briefing, share the trust FAQ, and test the anonymous reporting flow end-to-end.
Employees use the AI assistant and reporting in their own language. HR sees a case only when an employee reports; dashboards show protected aggregates.
Review wellbeing insights monthly, agree action plans, follow up, and produce board-ready evidence — measure, act, re-measure.
Raw conversations are encrypted and isolated at the database level — a guarantee enforced by the system, not a policy. HR only ever sees a case an employee chose to file. No emotion-recognition, no monitoring, no automated employment decisions.
Explore security & complianceRaw conversations are encrypted and isolated. HR only ever sees a case the employee chose to file.
No emotion-recognition, no keystroke or voice monitoring, no individual usage tracking shown to managers.
The AI never makes employment decisions. Outputs are advisory and reviewed by people.
Data minimisation, a retention period set by each company, audit logs and DSAR exports. The assistant says it is an AI whenever someone asks.
See which topics your people raise, in anonymous aggregates only, and act on them early.