1. AI systems
This document is a draft for founder and legal review. It is not legal advice and does not create a commitment until approved and published.
Bavio may use language, speech recognition, speech synthesis, extraction, summarization, and routing components depending on configuration. Provider availability and model behavior vary.
2. Accuracy and limitations
AI output can be wrong, incomplete, stale, ambiguous, or biased. A generated answer, Lead, insight, or transcript is not automatically a fact or professional opinion.
3. Human and AI disclosure
Customers must disclose AI interaction where required and provide a human path where appropriate. Do not represent an Agent as a person or claim that a caller reached a human when that is not true.
4. Automated actions
Actions and Workflows can create external effects. Execution evidence can show a request was attempted or acknowledged; it does not establish that a business, legal, financial, or factual outcome occurred.
5. Providers and data
Configured paths may use OpenAI, Cerebras, Groq, Sarvam, Deepgram, or ElevenLabs. The active provider, processing location, and retention terms depend on configuration and provider contracts.
6. High-risk uses
Do not use Bavio as the sole basis for emergency response, medical diagnosis, legal advice, financial advice, employment, housing, credit, insurance, or other high-impact decisions.
7. Voice and identity
Do not clone or imitate a real person’s voice without authorization, impersonate, spoof identity, or use generated speech for fraud or coercion.
8. Customer review
Customers are responsible for Agent instructions, Knowledge rights, caller notices, generated content review, escalation design, and the consequences of enabled automation.
9. Contact
AI and safety questions: [LEGAL_EMAIL]. Support: hello@bavio.in.