Pal, for every person
Everything that needs you, in one place.
Pal shows you what waits for your decision, how your flows are doing and what they cost. When it turns into real work, Pal hands it to the right AI agent.
Pal
Only you can see this chat
What’s waiting on me today?
- · Iris wants to refund €48.00 on order #20417
- · The returns flow waits at your checkpoint
- · One run failed: the shop didn’t answer
Your chats stay yours. Owners see the cost and the changes, never the conversation.
For everyone
Ask what waits, what failed and what it cost.
One conversation instead of five screens. Pal reads with your rights, so it only shows what you could open yourself.
What is waiting for you
Your approvals, the checkpoints in your flows and the runs that failed, in one answer. “What’s waiting on me today?”
How your flows are doing
What is running now, what starts next and what went wrong. “How is the returns flow doing?”
What it costs
The model runs of the past days and what they cost, by AI agent and by model. “What did Support cost this week?”
What happened
How the customer conversations on your channels ended, and what the audit trail says. “Who approved the refund on #20417?”
Hand-off
Real work goes to an AI agent.
Pal shows and hands over; your AI agents do the work in your systems. When a question turns into work, Pal passes it to the AI agent that can do it, context included.
You ask
“The customer of order #20417 wants a refund. Can someone take this?”
Pal proposes the hand-off
One card. You confirm it, and Iris takes the conversation with all of its context.
Iris takes over
Iris checks the order and prepares the refund. Anything that changes something waits for your approval.
Build mode
Pal helps build your AI team, if your role allows it.
People who manage your AI agents get a second mode. Describe what you need, and Pal proposes the AI agent, the AI team and a first draft of the flow, each as a card you approve.
We get a lot of webshop questions about returns. Can you set someone up for that?
Create an AI agent
- Name
- Sanne · webshop questions
- Team
- Support · reports to Iris
- May use
- Your shop (read orders) · Returns policy
- Asks you first
- Replies to customers
Trust
Private for you, accountable to the organisation.
People only use an assistant they can trust with half-finished thoughts. Owners only allow one they can account for. Pal is built for both.
Chats are private
Only the person who writes them can read them.
Cost per person
Owners see what each person’s Pal costs, within your AI budget.
Every change on record
Whatever gets built or handed over lands in the audit trail.
On or off per person
You decide who gets a Pal, one seat at a time.
Your model, your key
Your organisation’s routing policy picks the model, on your own key.
Hosted in the Netherlands
Like the rest of Infinized.
Where it stops
- Pal shows, starts and hands over. It never works in your systems itself: the AI agent that can do the work does it, under the same approvals as everywhere else.
Questions
What people ask first.
Read on in Learn
Working with AI agents: chat, approvals and PalIs Pal one of my AI agents?
No. AI agents do your organisation’s work in your systems. Pal helps one person, and passes real work on to the AI agents.
Who can read my chats?
Only you. Owners see what it costs and what it changed, not what you wrote.
Which model does it use?
The one your organisation’s routing policy allows, on your own key.
Can we switch it off for someone?
Yes. Owners and admins switch Pal on or off per person.
Give everyone a Pal.
We start by listening to your question. Then we show you a demo that fits your organisation.