No license per employee, no user fees, no subscription — with 2,000 people the maths works out differently than with 20.
We deliberately do not name a saving in percent: what you save depends on what you pay per head today.
Administration with all rights, supervisor without deleting and settings. Schedulers without a further role may read, record assignments, absences and notes, and use Klacksy.
You define per user who sees which groups. Schedulers then only see the employees of their groups; the administration sees everything. Employees without a group remain visible to everyone.
Sign-in via LDAP or Active Directory as well as via OpenID Connect and OAuth 2.0. You take over employees by synchronising with LDAP or Active Directory, which you trigger in the settings.
At period closing, Klacks creates the payroll export as a file — among others for DATEV and Abacus, as well as CSV or Excel.
You import employees from an Excel or CSV file; a preview shows what is missing before anything is saved.
Every record stores who created it, last changed it or deleted it, and when. Period closings, reopenings with a stated reason and daily approvals are kept in a separate log.
Assignments can be confirmed, approved and closed; each level locks editing.
The installation package comes with an update service that applies signed new versions of server and interface together.
Klacks was developed and tested with a dataset of 5,000 employees.
Klacks comes with its own MCP server. MCP (Model Context Protocol) is an open standard through which AI agents work with software. Any agent that supports MCP — such as Claude, Cursor or your own agents — can use it to work directly with Klacks: with the tools Klacksy uses as well, and only with the rights of the connected user.
Forming groups, setting up scenarios, recording entries: by sentence, with a preview and only after your yes. The rule-based planning engine calculates the plan. It takes the stored rules into account – rest times, qualifications, hour limits – and shows violations before anything takes effect. What only planners know – who can be reached at night, who cannot be asked for a shift right now – is not in the system. That is why you decide.
Klacks provides Klacksy with its capabilities. You decide which AI model sits behind it: from a provider of your choice or on your own server. Klacks helps you find a suitable model: a built-in test shows whether it works for Klacksy.
The same capabilities are open to your own AI agents too: any agent that supports the open standard MCP can use them — only with the rights of the connected user.
The source code is public, the Playground is open: your IT team can check Klacks before we talk to each other.