Key Concepts
The vocabulary you will meet across Astervis — telephony, AI, and platform terms in one place.
Astervis sits on top of your Asterisk / FreePBX server, replicates every call into a local TimescaleDB store, and layers analytics and AI on top. That model brings a specific vocabulary with it. This page is a glossary you can skim once and refer back to — each term gets a plain definition and a link to where it's covered in depth.
Astervis is on-premise only. Your call data and recordings never leave your server. The single exception is AI analysis, which is optional — when enabled, only the audio of a finished call is sent to Google Gemini. See AI Call Analytics for the full picture.
Telephony concepts
These describe the raw call activity Astervis ingests from your PBX. If you run a call center, most of these will already be familiar — they're defined here in Astervis terms.
AI concepts
These terms come from the optional AI pipeline. When a call with a recording lands, a single multimodal Gemini call (on Google Vertex AI) does transcription, speaker separation, analysis, and — if configured — rubric QA, all at once. The results are written onto the call. Each term below is one part of that output.
Every AI output is optional and metered. The analytics platform works fully without AI; turning it on simply adds these fields to each call.
Platform concepts
These govern access, billing, and integrations — the plumbing around the product.
Where to go next
Last updated on