Inbound AI phone agents
Each business line gets its own voice agent. The AI answers, talks, extracts data, texts after the call, and can hand off to a human.
One agent per number
Phone number, agent ID, voice, language, first message, and script — independently configured and toggled active or inactive.
Live transfer to a person
Transfer number, sensitivity, custom trigger phrases, announcement, and retries so the handoff sounds deliberate.
Call duration cap
Set a maximum conversation length in minutes so calls cannot run away.
Structured data from the conversation
Define extraction fields with an instruction and type (string, boolean, and more). After the call, Grepiv fills them from the transcript.
Pre-call CRM enrichment
An append webhook runs before the agent speaks. Template the body with caller_id, to_number, and other call variables so the AI already knows who is calling.
Conditional lead creation
Create a CRM lead only when extracted variables match rules — equals, contains, exists, numeric compares, AND / OR.
Post-call AI texting
If a condition matches, send an SMS after the call using a message template and a chosen from-number.
Post-call webhook workflows
Visual workflow builder: chain HTTP webhooks with conditions, headers, body templates, and a test / dry-run from the editor.
Mid-call tools on the number
Arm account-level tools so this inbound line can fire APIs or inject knowledge during the call.