Security

Connecting an AI system to your support tools means giving it access to real customer data. Here is what it can see, what happens to that data, and what stays under your control.

Read-only access. No refunds, edits or exports.
Own orders only. Each customer only reaches their own orders, checked on every request.
Rules enforced in code. A malicious message cannot talk the agent out of them.
Minimal data. Only the fields an answer needs are read.
No training on your data. Your customers' data never trains a model.
Full audit trail. Every data lookup is logged.
Instant off switch. One switch hands everything back to your existing setup, at any time.
Gradual rollout. Live on a share of your traffic first, then extended.

Connecting your tools

Least-privilege access
Read-only scopes to the data it needs: orders, delivery, catalogue. The key you issue can look things up and nothing else.
Encrypted, and under your control
Credentials are held as encrypted secrets and customer data is encrypted in transit. The access you grant stays yours to change or withdraw at any time.
Your data, and how long it stays
Used only to run and improve your system, isolated per client. You set the retention window, and everything is deleted on request.
Where it runs
On infrastructure Gem CX designs and operates, encrypted in transit, with a separate environment per brand.

GDPR

Lawful basis and roles
You remain the data controller. Gem CX acts as your processor, under a data processing agreement (DPA).
Data minimisation
The agent reads the fields an answer needs, scoped to the customer asking, and keeps no more than that.
Subject requests
Access, correction and erasure requests are answered from the logs, per conversation or per customer.
One accountable partner
Gem CX designs, runs and answers for the system end to end, so there is no chain of vendors to trace when a privacy question comes up.

Integrations

The system sits inside your conversation channels. Answering a customer means calling your store and third-party APIs and reading what they return, mid-conversation, not at the end of it.

Customer

Back-end