Security
Security and continuity
Ambit System holds HR records, candidates, contracts and company documents, so security is built into every module - from signing in to the history of changes. Below is everything that works today, described the way it works.
Sign-in and access
Who can get in, how they sign in and what they see once inside.
Two-factor sign-in
Codes from an app (TOTP) and recovery codes. The administrator can require them from administrators or from everyone. After a series of failed attempts the account is locked for a while.
Single sign-on
An organization's administrator connects the company's own sign-in - Microsoft 365 (Entra ID), Google Workspace, Keycloak or another OpenID Connect provider - without our help. People sign in at the organization's own link with their work account; only people the administrator added get in, or - if the administrator allows it - new people from the company's e-mail domains. The client secret is stored encrypted and never shown. A session from SSO lasts about half a day, so an account switched off at the provider soon stops working.
Sign-in with a code by e-mail
An administrator can turn on signing in with a one-time code sent by e-mail instead of a password. The code has 8 digits and works for a few minutes; if the account uses app codes, the system still asks for one.
Lockout after failed attempts
After a series of wrong passwords or codes the account is locked for a while, and the number of sign-in and password reset attempts from one address is limited. A correct password does not reset the counter until the code is right too.
Password rules
Every new password is checked against the password policy: a minimum length, a digit and a special character. A password reset works through a one-time link that expires soon.
Switching an account off works at once
An administrator switches an account off with one click: every session, trusted browser and token ends, and the account can no longer sign in any way. The person's data and history stay.
Permissions by role
Everyone sees and changes only what their role in the organization allows.
Data and privacy
Where the data is, who sees it and how we protect it.
Every organization's data is kept apart
Every record, query and file belongs to one organization. Another organization's file simply does not exist for you - the system does not even confirm that it is there.
Secrets only encrypted
KSeF tokens and certificates, the SSO client secret and the two-factor keys are stored only encrypted. A KSeF token is kept only once KSeF accepts it. The panel never shows them - not to you, not to us - and they are not exported.
PESEL and account numbers hidden in the history
The event log shows a PESEL number and a bank account number only by their last characters, e.g. ••••••••1234. The change is recorded, but the full value cannot be read from the history.
Candidates' data under the GDPR
A candidate's consents are stored together with the wording of the clause they accepted. On request you erase a person's data from every place in the system - files and error logs included - or prepare an extract of their data.
CVs checked for viruses
Every CV, also one sent through the job board's form, must be a real PDF and pass an antivirus scan. A file that cannot be checked is not stored.
Export of all data
The organization's administrator can ask, at any time, for an export of everything the organization holds: the history of changes, the current records and the files, in an open format (JSON). We prepare it in the background and e-mail a link to a page where the administrator downloads the ZIP file for a few days. Passwords, tokens and keys are never exported.
The trail of changes
Who changed what and when - with no way to rewrite the history.
The event log
Companies, opportunities, legal entities, job descriptions, job posts, candidates, applications, projects, workers, assignments and HR cases have an event log: who changed what and when, field by field. Every change is stored for good, and the stored history is never overwritten.
Support is visible
When our support team enters your organization as one of its members, the event log marks every such change clearly.
Continuity
Where the system runs, how we keep it running and what happens when something goes wrong.
Data in the European Union
The system and its data run in UpCloud's data centre in Warsaw.
Service status in public
Every minute we check from outside that the application, the API, the job board and the demo answer. The result and its history are on status.ambitsys.eu, and an outage raises an alarm for us.
SLA: a critical failure fixed within 3 hours
When the system is unavailable or a key function does not work for everybody, we restore it within 3 hours of detecting it or of your report (8:00-22:00 Polish time); one from the night is fixed in the morning, by 11:00. Every day, weekends and holidays included. It is in the terms (section 10).
A daily database backup
Every night we back up the whole database and keep the last few copies.