AI Use & Transparency

1. Our principle

AuraGrid builds a governed AI operating layer. AI helps with intake, qualification, drafting, routing, reminders and oversight, but permissions, approvals and verification stay in view and a person remains in charge of policy. An action is reported as done only after the connected system confirms it.

2. Where AI is used

• Understanding the intent of an inbound message or inquiry and producing a structured summary.
• Drafting replies, follow-ups and reminders from approved templates and information.
• Suggesting a route, owner, priority or next step according to rules the customer defines.
• Comparing products and preparing an order handoff from a catalog the customer connects.
• Summarizing status and flagging exceptions for human review.

3. Human oversight

• Customers define policies, approval thresholds and named escalation owners.
• Uncertain, sensitive, conflicting or out-of-scope cases are held for a person instead of being answered or acted on automatically.
• External messages and writes to connected systems require configured permissions and can require approval.
• Unverified outcomes remain pending or failed and are never shown as success.
• Decisions and actions are recorded so that they can be reviewed.

4. Limits and risks

AI can produce output that sounds confident but is incomplete, outdated, biased or wrong. It can misunderstand language, accents, dialects and context, and it can behave differently across languages. Translated text may contain errors. Do not rely on AI output as legal, financial, medical, tax or other professional advice.

AuraGrid does not guarantee any sales result, collection result or return on investment from using AI features.

5. High-stakes uses

The Services are not designed to make final decisions about a person's employment, credit, insurance, housing, education, healthcare, legal rights or access to essential services. Customers must not use the Services as the sole basis for such decisions. If a customer uses AI output as one input, the customer is responsible for human review, bias testing, notices, appeal routes and compliance with laws on automated decision-making and discrimination that apply to it.

6. Telling people they are dealing with AI

People should know when they are interacting with an automated system. The Reception product is designed to identify itself as automated where required by law, by the channel, or by the customer's policy. Customers must keep this setting on wherever the law requires it, including chatbot and bot-disclosure laws in the United States (such as California's bot disclosure law), the transparency obligations of the European Union AI Act, and similar rules elsewhere, and must not instruct the Services to pretend to be a human.

Calls that use synthetic or AI-generated voices are subject to special consent rules in several countries. See the Messaging & Calling Compliance policy.

7. Your data and AI

We do not use Customer Content to train general-purpose AI models for other parties. Third-party model providers are listed on the Sub-processors page when they process customer data. Customers should not submit data they are not allowed to share with those providers.

8. AI-generated content

Customers are responsible for how generated content is used. If a customer publishes synthetic images, audio or video, or text on matters of public interest, the customer must label it where the law requires. Do not use the Services to create deceptive deepfakes or to impersonate real people.

9. Reporting a concern

If you think AI output was harmful, biased, inaccurate or misleading, or you were not told you were dealing with an automated system, tell us. We will review it and work with the customer concerned.

AI concerns

Asia, including Bangladesh

Company

Postal address