Jonas connects Personio and employment contracts with employment law. He warns you in good time about probation ends, fixed-term contracts and notice periods – with a citation.
The data sits in Personio, the rules in the contract.
Start dates, fixed terms, special clauses in the employment contract and the law: anyone who wants to keep a deadline reliably in view has to bring four sources together.
The question
You ask Jonas – in one sentence.
Instead of building reports, you simply ask. Jonas shows what gets checked – step by step, across all sources.
The evidenced answer
Every finding comes with its citation.
Every statement names the source it comes from. Whatever can't be evidenced, the agent marks as open.
The next step
Jonas prepares – you approve.
Jonas prepares the next step. It is only carried out once you approve it.
Hi, I'm Jonas – your AI agent for HR & employment law. I keep an eye on your systems and speak up when something stands out.
I noticed four signals today – each inconspicuous on its own:
Signals · today
PersonioNew hires and fixed terms
SharePointEmployment contracts
OutlookManagers' calendars
c:node GraphBGB § 622 · TzBfG § 14
@Jonas, which deadline is critical next?
I'm checking all sources:
Read start dates from Personio
Checked employment contracts
Checked fixed terms against the TzBfG (Part-Time and Fixed-Term Employment Act)
Checked managers' calendars
J. Klein's probation ends on 14 Oct – the manager is on holiday from 10 Oct.
Termination with shortened notice only possible until 14 Oct
M. Klein's fixed-term contract expires on 31 Oct
Other deadlines in October not critical
Personio · Start dateEmployment contract J. Klein § 2BGB § 622 (3)TzBfG § 14
I've prepared the next step:
Draft · Probation review meeting
Invitation to the team lead and J. Klein for 8 Oct – with a note on the deadline.
ApproveEditNothing goes out without approval
Example case with sample data.
c:node Research · HR & Legal
Why HR deadlines slip through — and how traceable AI keeps them reliably visible.
Reading time ~5 minBasis 6 sources, verifiedTopic HR deadlines & contractsApproach source-backed · on-prem possible
Fig. 0 · Probation, fixed term, notice period — every deadline has a basis in the contract. Whoever overlooks it decides too late.
Summary
HR deadlines — probation, fixed term, notice period, reference letter — announce themselves, but their basis sits in the contract, their date in the HR system and their consequence in the calendar. Calendar reminders are maintained by hand and full of gaps, general AI guesses. This article shows, with sources, why that is risky — and how c:node binds every deadline to its contract clause and the law and makes it visible early.
1 The problem: deadlines have a basis — that nobody looks up.
A fixed-term contract ends on 31 Dec, the notice period follows from §2 of the contract, the probation period expires — each of these deadlines has a concrete citation. But the date lives in the HR system, the clause in the contract PDF, the obligation to act in the calendar. For a deadline to land on the desk in time, someone has to connect these sources by hand. That's exactly where the expensive mistakes happen: not because the information is missing, but because it is unconnected.
2 Why calendars and generic AI don't solve this reliably.
Calendar reminders are only as good as the hand that maintains them — an overlooked fixed term creates no reminder. And an AI chatbot? The US standards institute NIST states that fabricated statements are inherent to the design of such models; the chatbot confidently names a wrong deadline together with a made-up justification.1 For deadlines with legal effect, a convincing wrong answer is especially dangerous. Even when the AI searches the real documents, the error rate only drops — it does not disappear.2
3 The approach: bind every deadline to its clause.
c:node connects the HR system, contracts, calendar and payroll and lets c:node reason over a knowledge graph — people, contracts, deadlines and rules as a network of nodes and edges. Research calls this combination of pattern recognition and firmly checked rules neuro-symbolic AI; it delivers answers that can be justified and traced.3 Via the graph, c:node thinks several steps ahead — person → contract → clause §2 → next deadline — which simple search cannot do.4 Every deadline it names carries its citation with it.
:nodeGraph— HR deadlineslive
Source (API)EntityNext deadlineRule / evidence
Nodes active · edges fire in real timeisolated per tenant · grows with every contract
Fig. 1 · People, contracts, clauses and deadline rules connect into the next due deadline. c:node fires the path — every step stays bound to its citation.
4 In practice: scattered data becomes a traceable deadline.
The same question that today costs a manual review of all contracts, c:node answers in seconds — and discloses every statement:
Which HR deadline is critical next?
M. Klein — fixed term ends 31 Dec. Per §2 of the contract the notice period is three months, so the last day to act is 14 Oct. Two more fixed-term contracts expire in Q1. Recommendation: decide on an extension by 14 Oct, otherwise the contract automatically becomes permanent.
PersonioContract §2CalendarDeadline rule
5 Why this also holds up legally and organisationally.
Because c:node backs every deadline with clause and history, every decision can be traced afterwards — towards the works council, lawyers or court — without you having to disclose the inner workings of the model.5 And because c:node runs on-prem or in the EU cloud, sensitive personnel data stays in the EU or in your own data centre. Data protection is built in, not bolted on.
In short: Deadlines don't slip through because they're unknown, but because their basis sits unconnected. c:node connects it — and makes every deadline visible early, traceably and auditably.
Sources
NIST, Artificial Intelligence Risk Management Framework: Generative AI Profile (NIST AI 600-1), 2024 — fabricated statements are inherent to the design. nvlpubs.nist.gov…
MEGA-RAG (Multi-Evidence RAG) — retrieval reduces fabricated statements by more than 40 %, but does not eliminate them. PMC12540348, 2025. ncbi.nlm.nih.gov…
Hitzler et al. (eds.), Handbook on Neuro-symbolic AI and Knowledge Graphs, IOS Press, 2025. iospress.nl…
GraphRAG-Bench — graph-based AI outperforms plain text search in multi-step reasoning. arXiv:2506.02404, 2025. arxiv.org…
European Union — EU AI Act, High-level Summary: logging, documentation, human oversight for sensitive AI. artificialintelligenceact.eu…
The sources support the principles (limits of generative AI, retrieval reduces errors, neuro-symbolic approach, multi-hop superiority, EU AI Act requirements). Names, dates and figures in the live graph and the sample dialogue are simplified illustrations, not real cases.
Get started
Three ways to c:node.
Same product, same evidence – you choose where it runs.
PersonioNew hires and fixed terms
SharePointEmployment contracts