Glossary
The voice AI glossary for people buying AI calling.
This glossary defines 40 terms you meet when evaluating AI voice agents: the speech technology behind them, the telephony that carries the calls, the metrics that show whether they work, and the Indian and global rules that apply to calling people. Each entry gives a short answer first, then how it works, a worked example or common mistake, and where it shows up at Vaami.
Technology
The speech, language and analytics building blocks of a voice agent.
- AI Calling AgentAn AI calling agent, often called an AI telecaller in India, is a voice agent that places outbound calls on its own, for tasks such as lead...
- AI Voice AgentAn AI voice agent is software that holds a spoken phone conversation in real time: it listens, understands intent, decides what to do, takes actions...
- Barge-in (Interruption Handling)Barge-in is a voice agent's ability to stop speaking the moment a caller starts talking, listen to the interruption, and respond to it, the way a...
- Code-SwitchingCode-switching is when a speaker alternates between two or more languages within a conversation or even a sentence. It is the normal way many Indian...
- Conversational AIConversational AI is technology that carries on multi-turn dialogue with people in natural language, by text or voice, tracking context across turns...
- Endpointing (Turn-Taking)Endpointing is how a voice agent decides that the caller has finished speaking and it is now its turn to respond. Getting it wrong means either...
- Hallucination GuardrailsHallucination guardrails are the controls that stop a voice agent from stating things that are false or unapproved: restricted knowledge sources,...
- HinglishHinglish is the informal blend of Hindi and English that millions of Indians speak and write, switching languages mid-sentence. It is a distinct...
- RAG (Retrieval-Augmented Generation) for Voice AgentsRetrieval-augmented generation (RAG) lets a voice agent look up passages from your approved documents at call time and answer from them, rather than...
- Speech AnalyticsSpeech analytics is software that transcribes recorded or live calls and analyses them at scale for keywords, sentiment, silence, talk ratio and...
- Speech-to-Text (ASR)Speech-to-text, also called automatic speech recognition (ASR), converts spoken audio into written text. In a voice agent it is the listening step...
- Text-to-Speech (TTS)Text-to-speech (TTS) converts written text into spoken audio. In a voice agent it is the speaking step, and its time-to-first-audio and naturalness...
- Voice AIVoice AI is the umbrella term for technologies that let machines understand and produce human speech: speech recognition, language understanding,...
- Voice IntelligenceVoice intelligence is the analysis layer that turns recorded phone conversations into structured data: transcripts, summaries, sentiment, topics,...
Telephony & routing
How calls are carried, dialled, routed and handed to people.
- BYOT (Bring Your Own Telephony)BYOT, bring your own telephony, means connecting your existing carrier or SIP trunk to a voice AI platform instead of buying phone minutes from the...
- Conversational IVRConversational IVR replaces keypad menus with open speech: callers say what they need in their own words and an AI understands, routes or resolves...
- CPaaS (Communications Platform as a Service)CPaaS is a cloud service that exposes calling, messaging and video through APIs so developers can add communications to applications without running...
- IVR (Interactive Voice Response)IVR, or interactive voice response, is the automated phone menu that greets callers and routes them by keypad presses or short spoken commands, such...
- Outbound Dialer (Predictive, Progressive, Power)An outbound dialer automatically places calls from a contact list. Power and progressive dialers call one number per free agent; predictive dialers...
- SIP TrunkingSIP trunking carries phone calls over the internet between a telephony provider and your systems using the Session Initiation Protocol. It is how a...
- Warm TransferA warm transfer hands a live call from an AI agent to a human along with context, such as a summary, so the person does not start over. A cold...
India compliance
TRAI, DLT, DND and data-protection rules for calling Indian numbers.
- 140 / 160 / 1600 Number SeriesIn India, TRAI directs businesses to use dedicated number series for commercial calls instead of ordinary mobile numbers: 140 for promotional calls,...
- DLT RegistrationDLT registration is the process in which a business registers its entity, sender IDs (headers) and content templates on a telecom operator's...
- DND Registry (NCPR)The DND registry is India's national list of customers who have opted out of unsolicited commercial calls and messages, either entirely or by...
- DPDP Act (India)The Digital Personal Data Protection Act, 2023 is India's data-protection law. Its Rules were notified in November 2025 and apply in phases, so...
- TCCCPRTCCCPR is TRAI's rulebook for commercial calls and messages in India. It sets out registration of senders and templates, customer preference choices,...
- TRAI (Telecom Regulatory Authority of India)TRAI is India's statutory telecom regulator. For voice AI buyers it matters because TRAI's regulations govern commercial calls and messages,...
Global compliance
Consent and recording rules that apply beyond India.
- Call Recording ConsentCall recording consent is the legal requirement to tell callers, and sometimes obtain their agreement, before a call is recorded or transcribed....
- TCPA (Telephone Consumer Protection Act)The TCPA is the US federal law restricting autodialed, prerecorded and artificial-voice calls and texts. Since a 2024 FCC ruling, AI-generated voices...
Metrics
The numbers used to judge a voice agent honestly.
- Average Handle Time (AHT)Average handle time (AHT) is the mean time spent on a customer call from pickup to finish, including talk time, hold time and after-call work. It is...
- Call DeflectionCall deflection is steering routine enquiries away from human agents toward self-service such as an AI voice agent, app or FAQ. The aim is to keep...
- Containment RateContainment rate is the share of calls an automated system finishes without handing them to a human. It measures how much the AI handled, but not...
- First Call Resolution (FCR)First call resolution (FCR) is the percentage of customer issues fully resolved on the first call, with no follow-up call or callback needed. It is...
- Latency (End-to-End Voice Latency)End-to-end voice latency is the delay between a caller finishing a sentence and hearing the agent begin to reply. It combines endpointing,...
- Word Error Rate (WER)Word error rate (WER) is the standard accuracy score for speech recognition: the number of substituted, deleted and inserted words divided by the...
Business
Commercial terms and use cases buyers search for.
- AI ReceptionistAn AI receptionist is a voice agent that answers a business's incoming calls, handles common requests such as bookings, questions and message-taking,...
- Lead Qualification (BANT)Lead qualification is checking whether an enquiry is worth a salesperson's time. BANT, a common framework, scores a lead on Budget, Authority, Need...
- No-Show RecoveryNo-show recovery is the practice of contacting people who missed or are likely to miss an appointment, by call or message, to confirm, reschedule or...
- Per-Second BillingPer-second billing charges for the exact seconds a call or AI session lasts, instead of rounding each call up to the next minute. It matters most for...
- White-Label Voice AIWhite-label voice AI is a voice agent platform that an agency or software company can resell under its own brand, with its own domain, dashboard and...