HR
An agent is deciding about a person. Who checks it, and is the record fair?
A hiring agent and a refund agent should never have the same permissions. LeafCraft gives HR agents narrow access, a named checker and a record of the work, not of people.
Read requests and check them against policy
None about a person. Prepares, never decides.
Passes every request to the line manager with the policy reference
HR business partner, then HR head
Never reads pay, health or performance data
HR audit agent. Never itself.
What's changing in HR
Agents decide about people
Screening, onboarding, leave and requests now pass through agents. Each one is a decision about a person.
Fairness must be shown, not claimed
If a candidate or an employee asks why, the answer must be a rule and a named person, not a model.
Employee data needs the narrowest access
An agent should see only what its role needs, and nothing it keeps as a lesson.
Two risks the layer removes
The agents you can't see, and the model that gets retired.
Agents you don't know about
In HR, agents arrive through vendor updates and team pilots. LeafCraft finds all of them, gives each an owner, and brings it under your rules or removes it.
A model that gets retired
An agent built straight on one model has to be rebuilt and re-approved when that model is retired or repriced. On LeafCraft you switch the model for that agent in config and keep everything else. Months and a budget, saved.
- The model is retired or repriced
- Prompts, tests and sign-offs rebuilt
- Months of work, a new approval cycle
- The history stays with the old tool
- Pick another model for that agent
- Switching is a config change
- Test it against your own record first
- Rules, limits, owner and memory stay
Why it matters now
In HR, the agents arrived before the rules did.
HR agents are useful because they are fast with the routine. They are risky because every decision is about a person. The layer makes sure a person owns every decision about a person, an agent sees only the data its role needs, personal details are masked before anything is kept, and the record shows the rule that was applied.
The core infrastructure
What LeafCraft runs for HR.
One team
Agents handle the routine within their limit. Every decision about a person is owned by a person. A different agent or person checks the work.
How it works02Your rules
Access set per role, with employee data limited to what each role needs. Escalation on sensitive cases to named HR leaders.
How it works03Your memory
Every request, decision and reason on the record, with personal details masked before anything is kept as a lesson.
How it works04Your budget
Each agent and task inside a cost limit set before it runs, so HR can plan AI spend like any other budget line.
How it worksAI and people, one team
Your HR team, with AI on the org chart.
The agent prepares the case. Your person makes the call and writes why. A separate audit agent checks the decision against your rules.
What changes for you
Checked the request against the leave policy.
Passed to the line manager.
Approved.
Checked the rule applied. Passed.
How it helps your people
Who it helps around HR.
Industries where it runs
Leaders who ask first
Questions teams ask
Is the record used to watch employees?
No. It is a record of the work: tasks, decisions and the reasons behind them. It is not surveillance of people.
Can personal details end up in agent memory?
When an agent saves a lesson, emails, phone numbers, dates and reference numbers are masked first.
Can we erase data on request?
Yes. One call deletes everything stored for a session, and the deletion itself is recorded with who asked for it.
Bring one HR process.
We'll map who does the work, who decides and who checks, set the cost limits and show you the record.