The agentic layer, as a product you can buy.
Receivables and Risk work without it — rule-based, as always. The Agentic Suite adds the preparation on top: classify, justify, propose. Releasing is still a person's job.
Hello — what can we get done for you today?
Six things that do not exist without a person.
None of them replaces an existing process. They are six additions at the places where somebody currently reads, sorts and compares lists — prepared by the suite, still decided and released by a person.
Check groups
A run is grouped by how much review it needs, computed deterministically and justified per case.
The reasoning
The route to a proposal is readable while it happens — not just the result.
Questions back
Where something is ambiguous the agent asks with clickable options instead of guessing.
Analyses on demand
Analyses of a customer, a run or a selection, with the figures from the database.
Plans
Several steps as one proposal, releasable step by step — and nothing runs unreleased.
Lists
A selection that outlives the session and can be used on in the interface.
Three things our AI explicitly does not do.
This is the part that decides a rollout: the suite changes nothing about the rules that applied before it.
The release
No tool grants a release. What gets written is released by a person — for limits as for reminders.
The permissions
An agent reads with the acting person's roles and data scopes, and nothing when in doubt.
The record
No second store appears. The agent works on the same record as the interface.
Consumption is stated, per result and per session.
An agent consumes compute, and that is a line nobody should estimate. Every plan, every analysis and every summary states its own consumption; above the composer a counter runs for the whole session.
This page carries no prices, because the scope depends on which products you run and how much of it actually happens. You are billed for what you use — you pay nothing for parts you do not need. The full price overview and the calculator are on the pricing page.
See prices and calculatorSession balance
Session running 36 minutes · 14 actions
Balance
This session analysed and annotated 5 customers carrying 18.58m EUR in overdue volume; the group was rated critical. The standout is Becker Industries SE (no. 64618837), where a dunning block over 4 open disputes is holding 4.29m EUR of overdue receivables. The run produced 3 letters, created and sent — 14 actions across 8 objects over 2 analyses.
Analyses in this session
Session results
5 customers analysed and annotated, 3 letters created and sent.
Enabled per company, not per click.
The suite is switched on for your company rather than booked per user. Which parts you use follows from which products you run — the agentic dunning run needs Receivables, the credit preparation needs Risk.
With Receivables
Check groups in the run, clarification proposals, analyses of runs and customers.
ReceivablesWith Risk
Prepared limit suggestions with their reasoning, analyses of classes, cover and securities.
RiskWith both
An agent that knows both sides — the reminder sees the limit, the limit decision sees the payment behaviour.
What wants settling before a rollout.
No. Receivables and Risk are rule-based and work without it. The suite starts where somebody currently reads, sorts and compares lists — it is an addition, not a prerequisite.
Read, classify, justify, propose and prepare. Write only through declared actions, and the consequential ones only after a release. There is no tool it could use to grant itself that release — that is architecture, not a setting.
Through the consumption stated per result and per session. This page names no prices, because the scope depends on the products you run and on actual usage; the mechanism for checking it is visible before an invoice arrives.
A language model is called to phrase a proposal — what is transmitted and what is not is set out with evidence on the data-security page. What holds here: no model training on your data, no figure from the model, no cache in the subagent stack.
With one area and one run. The suite is switched on per company, and because every proposal needs a release, the first month is an observation: you see what the agent proposes and keep deciding yourself.