Delivered with Tugi Tark
AI player support, delivered with TUGI
Payments, bonuses, verification and the state of a request already in progress, answered in the player's own language. Anything needing judgement reaches your team.
What TUGI answers for players
The questions that generate most of the volume, at any hour and in the language the player wrote in.
Payments
Where a deposit is, why a card was declined, when a withdrawal was approved and what happens next.
Bonuses
What a bonus requires, how far a player is through the wagering, and why one expired.
Verification
Which document is still outstanding, what a rejected upload needs, and what the current account state means.
A request already open
The state of a case the player has already raised, without asking them to describe it again.
10M+
real iGaming support tickets in the training set
Up to 80%
of player tickets positioned as resolvable
249+
languages covered by real-time translation
Figures published by Tugi Tark for TUGI AI, not measured by Scalara. Which workflows are switched on, and how far each one goes before it hands over, depends on your deployment and the integration scope agreed in discovery.
Escalation hands your team the case, not a transcript
When a conversation needs judgement it moves to a person with the player record attached: account state, compliance tier, verification status and everything already said.
The conversation travels with the case
Your agent picks up where the player left off instead of asking them to start again.
You decide what may be answered at all
The subjects the assistant is allowed to handle are yours to set, and your support policy stays yours.
Every action is recorded
Assisted and self-service actions write to the same audit trail as a staff action.
What the assistant handles, and what stays with your team
Handled by TUGI
Repetitive payment, bonus and verification questions
The state of a request already in progress
Out of hours cover, in the language the player wrote in
Routing a case to the right queue with its context
Stays with your team
Judgement calls and disputes
Withdrawal review and payout decisions
Paying out is the irreversible action in a casino, so a person makes that call.
Compliance decisions and escalated cases
The support policy itself
You set which subjects the assistant may answer and what it must hand over.
Languages, and what coverage actually means
The language figure above is published by Tugi Tark rather than measured by us. What it means in practice is narrower than a headline number, and more useful.
A player writes in the language they are comfortable in, the assistant answers in the same language, and your agent reads the case in yours. Nobody waits for a translator, and no player is answered in a language they did not choose.
Coverage is not the same as competence in every subject. Which workflows are enabled, how far each one goes before it hands over, and which languages you switch on are decided when the integration is scoped with you. We would rather set that expectation here than let a number on a slide do it.
The question operators ask first
No, and we would not sell it that way. It handles repetitive volume, at any hour and in the player's own language, and escalates anything needing judgement to your team with the conversation attached.
What it changes is the shape of the queue. Your people spend their time on cases that genuinely need a person rather than on the same handful of questions in a dozen languages. You decide which subjects it may answer at all, and your support policy stays yours.
Talk to us about TUGI and what it would handle on your deployment
We will show you what the assistant answers today, what it hands over, and what the integration actually involves.
We reply within one working day.
