Built for
sensitive conversations

Controlled, overseen and secured.

Every decision stays with your team.


Four control layers on every message

ReMind is built only for responsible gambling conversations. Not general-purpose chat.

Defined scope

A specialised architecture with bounded models keeps every conversation on RG topics. ReMind’s instructions are kept separate from the player’s messages, which limits attempts to override them.

Output reviews

Every message passes automated safety and format checks before the player sees it. Messages flagged as harmful, off-topic or promoting gambling are stopped and sent for correction.

Correction & fallback

When a check flags an issue, the message is regenerated or the model is switched in real time. A safe default message is the last resort.

Human escalation

Critical cases, such as confirmed suicidal intent or AML concerns, are escalated to your specialists with full context. Crisis conversations continue until a human takes over.

These controls support safer, reviewable interactions.

They do not eliminate risk or replace your governance, policies, monitoring or regulatory obligations.

ReMind starts the conversation and documents it.

Your specialists decide on follow-up, monitoring and restrictions, under your own policies.

Your team keeps every decision

ReMind Does

Starts the conversation

Initial risk and affordability assessment

Raises alerts and escalates critical cases

Documents the intervention

Your team keeps

Ownership of RG policies and detection logic

Final account review and assessment

Final authority on every escalation

Decision on follow-up and restrictions

You configure what ReMind uses: your RG page, your RG tools, your escalation contacts.

Real-time alerts are returned through the API during the conversation, so your systems can bring in a specialist before it ends.

Everything documented. Nothing hidden.

FULL TRANSCRIPT

Every message from the player and ReMind, timestamped and stored.

RISK RATIONALE

The signals behind each assessment, backed by the player’s own words.

REAL-TIME ALERTS

Raised during the conversation, so your team can act before it ends.

STRUCTURED SUMMARY

Delivered as JSON, tailored to your team’s or monitoring provider’s format.

Designed to support your UKGC obligations

ReMind is designed to help you deliver and evidence customer interaction under SRCP 3.4.3

TIMELY INTERACTION

A behaviour-specific conversation starts within seconds of your API call, with no queue.

TAILORED MESSAGING

Reflects the play patterns you share, using motivational interviewing to explore intent.

EVIDENCE & EVALUATION

Transcripts and structured data support audits; pilots A/B test against your current process.

HUMAN OVERSIGHT

Severe harm indicators escalate directly to your specialists, with full context.

SR Code 3.4.3(13) asks operators to evaluate their approach, for example by trialling and measuring impact.

See how our pilot programme supports this.

Compliance with the UKGC’s Licence Conditions and Codes of Practice rests with the licensed operator.

No technology provider can be certified as compliant on its own.

ReMind runs in AWS EU region.

Your data is encrypted in transit and at rest, never used to train third-party models, and never shared with other operators.

EU-hosted, encrypted and never shared

Hosting

AWS, EU region (eu-central). US hosting available for US clients on request.

AI models

Anthropic models via Amazon Bedrock, in the same region.

No third-party training

Anthropic never receives or trains on your data.

One standard of protection

We review conversations to improve ReMind, for example to catch missed risks or improve player engagement. Every improvement reaches every operator, because player protection shouldn’t depend on where someone plays. Improvements are shared; your data never is.

Encryption

In transit and at rest, managed by AWS.

Access

Server-to-server calls with dedicated API keys.

Subprocessors

AWS only.

Data received

A player reference, age, language and country, plus any optional context you choose to share.

Engineering partner

Yktis supports on product architecture and robustness. It has no access to operator or player data.

Running due diligence?

We’ll answer your security, data and compliance questions directly.