This is Lynn.
The AI colleague in flynne: Lynn plans, warns and drafts so that your team can look after people instead of logistics. People do the judging. Always.
What Lynn takes on
We take AI where it genuinely relieves assessment operations.
Resolving disruptions
A psychologist drops out? Lynn already has the plan ready: a qualified stand-in, a new slot, notifications to everyone affected. You check and confirm with one click.
Suggesting staffing
The right psychologist for each event, by qualification, availability, location, workload and earlier assignments. Every suggestion comes with its reasons; you decide.
Drafting messages
Emails that fit the situation, to candidates and psychologists, in the right tone, as a draft in your outbox rather than a surprise in someone else’s inbox.
Seeing bottlenecks coming
Lynn spots it when the volume of registrations exceeds the psychologist days available: weeks before it gets tight, not on the assessment day.
Answering questions
“How many no-shows did we have in Q2 in Frankfurt?” Lynn answers from your real operational data, with a number, a table or a chart.
Briefing the day
Every morning: the appointments of the day, the risks, the open decisions, as a compact briefing before you have finished your first coffee.
How Lynn works
1
Lynn suggests
Out of availability, qualifications and operational data comes a concrete, reasoned proposal.
2
You check
The proposal shows its reasoning: who, why and with which alternatives. You change it or confirm it.
3
flynne logs it
Proposal, decision and execution land in the audit-proof activity log, just as with every human action.
What Lynn never does
Lynn does not judge people
No screening, no ranking, no suitability judgements. Assessment belongs to the psychologists, as a matter of principle rather than a roadmap item.
Lynn does not decide alone
Suggesting, preparing, reminding: yes. Assigning or cancelling on her own: no. The decision is always a person’s.
Lynn never works in the dark
Every action lands in the same audit-proof activity log as human actions. Your works council will like Lynn.
A morning with Lynn
The most common case in assessment operations: somebody drops out and the day runs anyway.
06:48
The cancellation arrives
A psychologist reports sick. Two procedures are planned with her today, fourteen candidates in total, six of them travelling more than a hundred kilometres.
06:48
Lynn works it out
Who has the qualification for this type of procedure, is free today, is not already double-booked and can reach the site in time? Of four possible people, two remain.
06:49
The proposal is ready
Two options, each with reasons. Option A moves one exercise by forty minutes and needs no stand-in. Option B brings a colleague from the neighbouring site who would arrive at 9.30, leaving the day unchanged but costing another procedure its second assessor.
07:15
A person decides
The coordinator picks option A, changes one small thing in the order and confirms. Lynn carried none of it out; up to this point everything was a proposal.
07:16
The messages go out
Fourteen candidates, three assessors, reception and the canteen receive the changed times. Lynn drafted the texts; a person read them and sent them. Everything is in the log: proposal, decision, execution, with time stamps and names.
What Lynn needs in order to be useful
The most common disappointment with AI features has no technical cause: it can only work with what is in the system.
Qualifications kept up to date
Lynn can only work with what is recorded. If a person has no qualification stored, she is suggested for no type of procedure; and anyone who records everything for everyone gets suggestions that do not hold professionally. The qualification matrix is the foundation, not an extra.
Recorded availability instead of shouted answers
Anyone who keeps their free times in their head and gives them on request is invisible to scheduling. Recorded as shift patterns or windows, they become capacity you can plan with. That is a change of habit and the real work during implementation.
Rooms and resources with properties
A room is not simply a room: size, equipment, accessibility, whether a group exercise works in it. Only with those details does Lynn spot conflicts that would otherwise surface on the day.
A few weeks of operation
Statements about bottlenecks and no-show patterns need a past. In the first weeks Lynn is therefore useful mainly in scheduling, not in prediction.
Where Lynn does not help
Because a page about AI without limits reads like a brochure, and is believed accordingly.
She does not replace coordination
An organisation that needs two full-time people for assessment organisation today will not need zero afterwards. What falls away is the searching: chains of phone calls, comparing calendars, chasing replies that never came. What remains are decisions and conversations.
She does not know your politics
The fact that a particular person should not work with a particular department sits in no data field. People have to correct suggestions like that, which is why the confirmation is not a formality but the whole point.
She does not negotiate
A proposal that costs a department head two days is not accepted because the arithmetic is good. People have that conversation; Lynn supplies the numbers for it.
She does not rescue a day in hindsight
If three assessors are missing at eight in the morning and there is no reserve, there is no solution, only a better cancellation. The value lies weeks earlier, in the planning, not in the rescue.
Common questions about Lynn
Where do the suggestions come from: does Lynn calculate or guess?
She calculates with what is recorded: qualifications, availability, workload, travel, deadlines. Every suggestion shows its reasoning and its alternatives. When the reasoning does not convince, something in the system is usually not maintained, and that is good news, because it can be fixed.
What happens when Lynn makes a poor suggestion?
Then it is not confirmed and nothing happens: a suggestion is not an action. If Lynn finds no solution, she says so rather than producing a poor one. If poor suggestions pile up, the cause may be the master data, missing qualifications or availability recorded too coarsely, or a rule that has to work differently at your house. Report such cases and we look at both sides.
Can we switch individual features off?
Yes, each one separately. Some organisations start with conflict detection only and add suggestions and draft messages later. The platform works fully without AI, only with more manual work.
Does Lynn rate candidates?
No: no screening, no ranking, no suitability statement, not even as a hint. That is a decision of principle, not a missing feature. Why we draw that line and how it can be checked is set out in full on our AI principles page.
What will our works council say?
From many completed works council processes in large groups: the point at which the conversation relaxes is the separation between organisation and judgement. We bring the reasoning and the documents; the points to check are set out in the AI principles.
The legal side, the classification under the EU AI Act and the points for the works council and data protection to check, is set out in our AI principles.
See Lynn in your own operation?
In 30 minutes we show you how Lynn plans, warns and drafts in your procedures.
We reply within one working day.