The words, without the marketing.

Agentic voice

Agentic voice describes voice systems that act autonomously toward a goal rather than only responding to what was said, with the emphasis on initiative and multi-step task completion.

AI receptionist

An AI receptionist is a voice agent doing the job a front-desk or switchboard person does, answering, greeting, handling routine questions, taking messages, booking and routing callers, which makes it a use case rather than a distinct technology.

Conversational AI

Conversational AI is technology that holds a back-and-forth dialogue with a person, in text or in speech, and it is the parent category that voice agents and AI chatbots both belong to.

Conversational IVR

Conversational IVR replaces the numbered phone menu with an open question, letting callers say what they need in their own words while keeping the same routing structure underneath, which makes it the transitional form between IVR and a true voice agent.

IVR

IVR, or Interactive Voice Response, is the automated phone menu that routes callers with keypad tones or a small set of recognized words, such as press one for sales and press two for support.

Multimodal agent

A multimodal agent is a system that interacts across more than one channel at the same time, most often speech, text and a screen, so it can hear a question, answer aloud, and show something on the page at once.

Voice agent

A voice agent is software you talk to out loud that hears you, works out what you are asking, decides what to do about it, and answers in speech.

Voice AI

Voice AI is the umbrella term for artificial intelligence applied to speech, covering recognition, synthesis, conversation, analytics, biometrics and generation, and it names a category rather than a single product.

Voice assistant

A voice assistant is a general-purpose spoken helper tied to a device or platform, such as those on phones and smart speakers, that works for the user across many tasks rather than for a single business.

Voice bot

A voice bot is an automated spoken system that follows an authored script or decision tree, handling the paths it was built for and falling back on anything nobody anticipated.

Voice user interface

A voice user interface (VUI) is the design discipline for systems people talk to, covering how a spoken interaction is structured, what the system says, and how it handles the moments that go wrong.

Voice widget

A voice widget is the embeddable front end of a website voice agent: the script tag that puts it on the page, the isolated interface a visitor taps, and the microphone and audio-transport layers behind it.

Website voice agent

A website voice agent is a voice agent that runs on a web page rather than a phone line, so a visitor can ask a question out loud and hear the site answer it from the information that business has published.

Automatic speech recognition

Automatic speech recognition (ASR), also called speech to text, is the technology that turns spoken audio into written text, and it is the first real step in almost every voice agent.

Barge-in

Barge-in is a voice agent stopping its own speech the instant a person starts talking over it, which is what makes the exchange feel like a conversation rather than a recording you have to sit through.

Custom vocabulary and keyword boosting

Custom vocabulary tells a speech recognition system which unusual words to expect, such as your product names and industry terms, and keyword boosting raises the chance the recognizer picks those words when the audio is ambiguous.

Diarization

Diarization is the process of working out who spoke when in a recording, separating a conversation into anonymous speaker turns without needing to know the identity of any speaker.

Echo cancellation

Echo cancellation stops a device from hearing its own voice, so that a voice agent speaking through a laptop speaker does not pick up its own audio through the microphone and mistake it for the visitor interrupting.

Endpointing

Endpointing is how a voice agent decides that the person has finished speaking and its own turn has begun, normally by timing the silence against a threshold set between 150 and 300 milliseconds.

Intonation

Intonation is the pattern of pitch across a phrase or sentence, the part of prosody that tells a listener whether they are hearing a question, a statement, or something not yet finished.

Neural TTS

Neural text to speech (neural TTS) uses neural networks to generate speech directly, replacing the older concatenative and parametric methods, and it is the reason synthetic voices stopped sounding robotic.

Noise suppression

Noise suppression removes background sound from captured audio so that speech is clearer, and it is what lets a voice agent work for someone in a cafe rather than only in a quiet room.

Phoneme

A phoneme is the smallest unit of sound that changes the meaning of a word in a language, so that "pat" and "bat" differ by exactly one, and it is the level at which a voice agent gets a name right or wrong.

Prosody

Prosody is the rhythm, stress, pace and pitch that run across a whole sentence rather than within a single sound, and it is most of what makes synthesized speech sound alive rather than flat.

Realtime API

A realtime API is an interface built for live, low-latency, bidirectional audio streaming rather than request-and-response calls, and it is what makes conversational voice possible without stitching separate recognition, reasoning and synthesis services together.

Sample rate

Sample rate is how many times per second audio is measured when it is digitized, and it sets a hard ceiling on which frequencies survive to reach the recognizer.

Speaker identification

Speaker identification works out who is speaking by matching a voice against a database of known voiceprints, answering who this is, which is a more legally sensitive question than who spoke when.

Speech-to-speech models

A speech-to-speech model is a single model that consumes audio and emits audio directly, with no transcription step in the middle, which buys latency and preserved tone at the cost of the text record a cascaded pipeline produces as a by-product.

SSML

SSML, the Speech Synthesis Markup Language, is a W3C standard for controlling how synthesized speech sounds by wrapping text in tags that set pauses, emphasis, pronunciation, speaking rate, and how numbers and codes are read.

Streaming against batch transcription

Streaming transcription emits words as they are spoken and revises them as more audio arrives, while batch transcription waits for the whole recording and returns a more accurate finished transcript.

Text to speech

Text to speech (TTS), also called speech synthesis, turns written text into spoken audio, and it is how a voice agent gets its voice.

Turn detection

Turn detection is working out whose turn it is to speak, and semantic turn detection does it by judging whether the sentence sounds finished rather than by timing the silence after it.

Voice activity detection

Voice activity detection (VAD) is the component at the very front of a voice agent that decides, frame by frame, whether the incoming audio contains speech or only background noise.

Voice cloning

Voice cloning builds a synthetic copy of one named person's voice out of a recording of them talking, and that copy can then be made to say things the person never said.

Wake word and hotword detection

A wake word is the phrase that activates a listening device, and hotword detection is the always-running, low-power on-device process that recognizes that one phrase and rejects everything else.

Character error rate

Character error rate (CER) is transcription accuracy measured over individual characters rather than whole words, using the same substitution, deletion and insertion formula as word error rate, and it is the right metric where word boundaries are unclear or where exact spelling matters.

End-to-end latency

End-to-end latency is the total elapsed time across a system from one named point to another, and in voice it usually means from the moment a visitor stops speaking to the moment they hear a reply, a number that means nothing until its start and stop points are stated.

Interruption handling latency

Interruption handling latency is how quickly a voice agent stops its own speech once a person starts talking over it, targeted at roughly 60 milliseconds and judged about an order of magnitude more strictly than the delay before it answers.

Jitter

Jitter is variation in the timing of arriving packets, and because audio playback needs a steady stream, a packet that arrives late is as damaging as one that never arrives at all, since the moment it was needed has already passed.

Mean opinion score

Mean opinion score (MOS) is a 1 to 5 quality rating for synthesized or transmitted audio, produced by asking human listeners to score samples and averaging their answers, where 5 is excellent.

Naturalness score

A naturalness score rates how human synthesized speech sounds on a roughly 1 to 5 scale averaged from listener ratings, and it has become a weaker way to choose a voice as leading systems cluster near the human ceiling of 4.5 to 4.7.

Packet loss

Packet loss is audio data that never arrives, and in voice the loss pattern, scattered or bursty, decides more than the raw percentage whether a listener hears a small artifact or a broken conversation.

Percentile latency

Percentile latency describes how latency is distributed rather than averaged, so a p95 figure means 95 percent of requests were faster than it and the slowest 5 percent, the turns a visitor actually remembers, were slower.

Real-time factor

Real-time factor (RTF) is how long a system takes to process audio divided by the duration of that audio, so a value below 1 is faster than real time and a value above 1 is falling behind, and it governs cost and concurrency rather than conversational responsiveness.

Round trip time

Round trip time (RTT) is how long a packet takes to reach a server and come back, the network floor beneath every other latency number, and the one component no amount of software optimization can push below.

Silence threshold

The silence threshold is the amount of trailing quiet a voice agent waits for before deciding the speaker has finished, usually set between 150 and 300 milliseconds, and it is the single most-felt setting in the pipeline because it is paid on every turn.

Time to first audio

Time to first audio (TTFA) is the end-to-end wall-clock delay from the moment a person stops speaking to the moment the reply becomes audible, and it decides more than any other number whether a voice agent feels alive.

Time to first byte

Time to first byte (TTFB) in voice AI is how quickly a speech synthesizer starts emitting audio after it receives text, which is the moment the listener's wait ends, and it shares its name with an unrelated web-performance metric that has to be qualified apart from it.

Time to first token

Time to first token (TTFT) is how long a language model takes to produce its first piece of output after a request, measuring the model alone rather than the full pipeline, which makes it one component of time to first audio and the most frequently misquoted latency number in voice AI.

Word error rate

Word error rate (WER) is the standard measure of transcription accuracy: the words a speech recognition system got wrong, counted as a percentage of the words that were actually said, where lower is better.

Agent orchestration

Agent orchestration is the layer that coordinates the parts of a voice agent: which model runs, which tools are called, in what order, and what happens when something fails.

Chunking

Chunking is splitting documents into smaller passages before they are indexed for retrieval, and where those splits fall largely decides whether the right answer can be found at all.

Context window

The context window is how much text a model can consider at once, counting both what it has been given and what it produces, and everything outside the window does not exist as far as the model is concerned.

Conversation memory

Session state is what a voice agent remembers within a single conversation and conversation memory is what it remembers across conversations, and the two are different problems carrying different risks.

Distillation

Distillation trains a smaller model to imitate a larger one, producing a compact model that performs close to the original on the tasks it was trained to imitate, at a fraction of the cost and latency.

Embeddings

An embedding is a list of numbers representing the meaning of a piece of text, arranged so that two passages about the same thing sit close together even when they share no words, which is what lets a system search by meaning instead of by keyword.

Fallback

A fallback is what happens when the primary path fails, covering everything from a voice agent not understanding a question to a model provider going offline, and its quality decides whether the system degrades or breaks.

Fine tuning

Fine tuning continues training a model on your own examples so it behaves the way you want, changing the behavior of the model rather than its knowledge of your current facts.

Function calling

Function calling is how a language model acts rather than only speaks: it emits a structured request for an action, and the application around it checks that request, runs it against a real system, and feeds the result back into the conversation.

Grounding

Grounding is the discipline of tying what an AI says to a verifiable source, so the answer can be traced to real material rather than to whatever the model absorbed during training.

Guardrails

Guardrails are the limits that keep a voice agent from saying or doing what it should not, whatever the visitor asked and whatever the system retrieved, and the thing that makes them guardrails is that they hold outside the model instead of being asked of it.

Hallucination

A hallucination is when an AI states something false with the same confidence it states something true, which is not a lie, because there is no intent, and not a bug, because the system is working as designed.

Human handoff

Human handoff is the act of moving a live conversation from an automated agent to a person, and escalation is the judgement that the move is needed, and together they are what a customer runs into whenever the automation could not finish the job.

Inference

Inference is running a trained model to produce an output, as distinct from training which builds the model, and in production it is where nearly all the cost and latency live.

Knowledge base

A knowledge base is the collection of material a voice agent answers from; in a RAG system it is the uploaded documents, and its quality sets the ceiling on answer quality.

Knowledge graph

A knowledge graph stores information as connected entities and relationships rather than as passages of text, so it can answer structured questions by traversing those connections rather than by retrieving a passage.

Large language model

A large language model is a neural network trained on vast amounts of text to predict the next token, and that single capability at scale produces systems that answer questions, follow instructions and hold conversations.

Prompt injection

Prompt injection is when text supplied to a model contains instructions that override or subvert its intended behavior, and it is the defining security problem of language model applications.

Reranking

Reranking is a second pass that reorders retrieved results by relevance using a slower and more accurate model than the initial search, and it is what fixes the near-misses fast retrieval gets wrong.

Retrieval-augmented generation

Retrieval-augmented generation, or RAG, is a technique that makes an AI look the answer up in specific documents you have given it before it replies, instead of answering from whatever it absorbed during training.

Semantic search

Semantic search finds results by meaning rather than by matching words, so a spoken question and the written answer can share no vocabulary at all and still be matched correctly.

Small language model

A small language model is a compact model with far fewer parameters than a frontier one, faster and cheaper and less capable, which for many narrow voice tasks is the right trade.

State machine against LLM-driven flows

A state machine follows an authored path with defined states and transitions, an LLM-driven flow lets the model decide what happens next, and the choice between them sets how predictable and how flexible a voice agent is.

System prompt

A system prompt is the standing instruction given to a model before any conversation begins, setting the role, the scope, the tone and the rules.

Temperature

Temperature is a sampling parameter that controls how varied a model output is, low producing consistent and predictable responses and high producing more varied ones, by reshaping the probability distribution over next tokens before selection.

Token

A token is the unit a language model reads and writes, usually a fragment of a word rather than a whole word, and cost, context window capacity and latency are all measured in it.

Vector database

A vector database holds embeddings and returns the nearest ones to a query fast enough that nobody waits, which makes it the search index that meaning-based retrieval runs on.

Audio codecs

An audio codec is the scheme that squeezes sound into a form a network can carry, and whichever one a voice system runs on settles how the audio sounds, what it costs in bandwidth, and how much of it a recognizer can actually transcribe.

Call routing

Call routing is the logic that sits between a call arriving and someone answering it, deciding which queue, agent, department or automated system the call should reach, and an AI voice agent typically joins that logic as one of its destinations rather than replacing it.

Cold transfer

A cold transfer, also called a blind transfer, passes a call to another destination with no context and no briefing, forcing the caller to re-establish their identity, problem and history; it is the default behavior whenever context passing is not deliberately built.

Concurrency

Concurrency counts the conversations a voice system can run at once, which makes it a question about capacity rather than about volume, and it governs what happens on the busiest day instead of the ordinary one.

Contact center platform

A contact center platform is the system that runs customer contact operations, meaning routing, queuing, agent management, recording, reporting and workforce scheduling, and it is the incumbent estate an AI voice agent has to work alongside rather than replace.

CPaaS

CPaaS, Communications Platform as a Service, is an API that lets software make calls, send messages and handle voice without the business negotiating with carriers or navigating numbering regulation in every market itself.

DTMF

DTMF, dual-tone multi-frequency, is the touch-tone scheme that turns every keypad press into its own pair of tones, and it remains the right way to collect the card numbers, PINs and reference codes that speech recognition reliably mangles.

Inbound vs outbound agents

An inbound voice agent answers contacts that come to the business while an outbound agent initiates contact, and although they share the same technology they differ so much in regulation, deliverability, intent and economics that they are substantially different businesses.

Number provisioning

Number provisioning is obtaining and configuring telephone numbers for a service, and because numbers are regulated resources allocated through national numbering plans, it is geography-specific and slower than most software teams expect.

PSTN

The PSTN, or Public Switched Telephone Network, is the worldwide arrangement of carriers, numbering plans and interconnect deals that lets any phone number dial any other, and the narrowband audio it carries holds measurably less information than a browser microphone captures.

SIP and SIP trunking

SIP, the Session Initiation Protocol, is what phone systems use to open, manage and close voice calls across internet networks, and SIP trunking is the service you buy to reach the outside telephone network through an internet connection rather than through physical lines into the building.

Softphone

A softphone is telephone software running on a computer or mobile device in place of a physical handset, and in voice AI it matters as the surface where a human agent receives an escalation and where the AI conversation context has to appear.

STIR/SHAKEN and spam-likely labeling

STIR/SHAKEN is a framework that lets carriers cryptographically verify a call really originates from the number it claims, while spam-likely labeling is a separate and proprietary scoring layer that decides whether a warning appears on the recipient screen.

Telephony provider and carrier

A carrier operates telephone network infrastructure and terminates calls, while a telephony provider is whoever a business buys voice service from, which may be a carrier or a reseller sitting above one, and the route between them determines cost, audio quality and attestation.

Warm transfer

A warm transfer passes a conversation to a person along with everything already known about it, so the receiving human arrives informed and the customer does not have to repeat themselves; the defining property is the completeness of the context, not the speed of the handoff.

WebRTC

WebRTC is a set of open browser standards for capturing, processing and transporting live audio and video in real time, and it is what lets a website hold a spoken conversation with no plugin, no app and no phone call.

Abandonment rate

Abandonment rate is the share of customers who give up before their issue is handled, whether by hanging up in a phone queue or closing the browser tab, and it is the correction that keeps a containment figure honest.

After-call work

After-call work is everything an agent does after the customer leaves, from writing notes to coding the disposition, updating systems and creating follow-up tasks, and it counts as part of average handle time even though planning routinely ignores it.

Agent assist

Agent assist is AI that helps a human agent during a live interaction by surfacing information, suggesting responses and handling documentation, rather than replacing the agent.

Average handle time

Average handle time is the mean duration of a customer interaction, including talk time, hold time and the after-call work done once the customer has gone; it is a cost driver, not a measure of quality.

Call deflection

Call deflection is the practice of moving a customer contact out of an expensive channel, usually the phone, into a cheaper one such as self-service, and it counts the contact that never arrived rather than the problem that got solved.

Containment rate

Containment rate is the share of conversations an automated system finishes on its own without handing the person to a human, calculated as conversations resolved without human involvement divided by total conversations handled.

Conversation completion rate

Conversation completion rate is the share of conversations that reached a natural ending rather than being cut short, a technical health metric that measures whether interactions finished, not whether they succeeded, and it is frequently mistaken for a quality metric.

Cost per call

Cost per call is the total cost of handling one customer interaction, including agent labor, technology and overhead, and it is the headline figure in every automation business case even though an honest one has to include the interactions the automation handles badly.

Cost per minute

Cost per minute is the unit price of voice processing, covering speech recognition, language model inference, speech synthesis and, in telephony, carrier charges, and it is how voice platforms bill and how voice unit economics are modeled.

Cost per resolution

Cost per resolution is the total cost of actually solving a customer problem across every attempt and every channel it took, divided by resolved issues, and it is the honest version of cost per call because it counts the failed attempts, escalations and repeat contacts that cost per call conceals.

CSAT

CSAT, or customer satisfaction score, is the percentage of customers who report being satisfied with a specific interaction, measured by a single question asked immediately afterward and answered only by the self-selected minority who respond.

Customer effort score

Customer effort score measures how much work a customer had to do to get their issue resolved, usually with a single question asked right after an interaction, and lower effort predicts loyalty better than higher satisfaction does.

Escalation rate

Escalation rate is the share of interactions an automated system hands to a human, the inverse of containment, and it is a better-behaved metric than its reputation suggests because correct escalation is a good outcome rather than a failure.

First contact resolution

First contact resolution is the share of customer issues fully resolved in a single interaction, with no callback, no follow-up and no second attempt, and it only holds up when it is measured across every channel rather than within one.

Intent recognition rate

Intent recognition rate is how often a system correctly sorts a customer utterance into one of a predefined set of intents, a legacy accuracy metric that stops being meaningful in generative systems because they have no fixed intent set to classify against.

Net promoter score

Net promoter score is a relationship metric that subtracts the percentage of detractors, who score zero to six, from the percentage of promoters, who score nine or ten, on a zero to ten recommendation scale, producing a number between minus one hundred and plus one hundred.

Occupancy

Occupancy is the percentage of logged-in time that agents spend actively handling interactions rather than waiting for the next one; it measures how intensively staff are used and has a sustainable ceiling above which it drives burnout and attrition.

Resolution rate

Resolution rate is the share of interactions where the customer issue was actually solved, which is a different question from whether the conversation ended without a human, and it is the metric containment is routinely mistaken for.

Service level

Service level is the percentage of contacts answered within a target time, conventionally written as two numbers such as 80/20, meaning 80 percent of contacts answered within 20 seconds.

Task completion rate

Task completion rate is the share of interactions where the customer accomplished the specific thing they set out to do, an outcome metric that suits generative voice agents because it needs no intent taxonomy, only a definition of what completion means for each task.

AI search

AI search means search interfaces that generate a composed answer rather than returning a list of links, so the user reads a synthesized response, sometimes with citations, instead of clicking through to the sources it was built from.

Answer engine optimization

Answer engine optimization is the practice of writing and structuring content so an AI answer engine picks it up and cites it, which means competing for a mention inside someone else's answer rather than for a ranking position on a results page.

Bounce rate

Bounce rate is the percentage of visitors who arrive and leave without any further interaction, and it is one of the most cited and least reliable web metrics because it counts a satisfied departure and a wrong-page exit as the same event.

Citation rate

Citation rate is how often a source is referenced in AI-generated answers, the metric that would replace ranking in a generative search world, and it is a sampled proxy for a non-deterministic behavior rather than a standardized figure that can be reported with the confidence of a ranking.

Click-to-call

Click-to-call lets a website visitor request a phone conversation with one click, usually by triggering a callback rather than making them dial, and it is the older bridge between a website and a live voice conversation.

Conversion rate optimization

Conversion rate optimization is the practice of increasing the share of visitors who take the action you want, by removing the reasons they do not convert rather than by buying more traffic, so every improvement applies to visitors you have already paid to acquire.

Crawlability

Crawlability is whether automated systems can access, read and understand a website content, and if a crawler cannot reach a page nothing else about that page matters, because content that cannot be reached cannot be cited, ranked, retrieved or summarized.

Dwell time

Dwell time is how long a visitor spends on a page or site before leaving, used as a proxy for whether content held attention, though longer presence can equally mean engagement, confusion, or a background tab nobody is watching.

Exit intent

Exit intent is a signal that a visitor is about to leave, most commonly detected by cursor movement toward the browser controls, and it triggers a last-moment intervention, though it is a probabilistic signal with substantial false positives that barely works on mobile.

Generative engine optimization

Generative engine optimization is the practice of improving a brand presence and citation within AI-generated answers, and it is an emerging discipline whose boundary with answer engine optimization is not settled.

Intent-triggered activation

Intent-triggered activation is a tool appearing at a judged moment based on visitor behavior rather than waiting to be clicked, so the system decides when to offer itself, and its entire value rests on the quality of that timing judgment.

Lead capture

Lead capture is collecting the contact details a visitor gives so a business can follow up, and it is the conversion event for anything that does not complete online; collected in a conversation, the details arrive with the context of what the visitor wanted, which is what makes a lead strong rather than merely present.

LLM SEO

LLM SEO is a third name for the same emerging practice that answer engine optimization and generative engine optimization also describe: being the source AI systems draw on when they answer questions.

llms.txt

llms.txt is a proposed convention for a markdown file at a site root that presents content in a form language models can consume easily, and it is an emerging proposal rather than a ratified standard that AI systems reliably consume.

On-site search

On-site search is the search box on a website, used by visitors to find something they could not locate through navigation, and its logs are the most direct available statement of what visitors want and cannot find.

Post-search web

The post-search web is a framing that describes a shift in how people find information online, away from searching and clicking through results and toward asking and receiving answers directly, and it is a positioning frame rather than an established industry term.

Schema markup

Schema markup is structured data added to a web page that tells machines what the content means, turning a price on a page into an identified price and an address into an identified address.

Scroll depth

Scroll depth is how far down a page a visitor scrolls, usually measured as a percentage of page height or in milestones, and it is the most reliable of the common engagement signals because it takes a deliberate action that a background tab cannot fake.

Session duration

Session duration is how long an entire visit lasts across all pages, before the visitor leaves or the session times out, and it is a weak quality proxy because efficient answer-finding produces short sessions while confusion produces long ones.

Site search abandonment

Site search abandonment is a visitor searching a website, seeing the results, and leaving without clicking anything, which is a search that ran and failed at the highest-intent moment on the site.

Voice commerce

Voice commerce is buying and selling through spoken interaction, from asking about products to completing a purchase, though its durable value lies in the pre-purchase conversation that removes an obstacle to a sale rather than in voice checkout itself.

Voice search

Voice search is speaking a query to a search engine or assistant instead of typing it, which is a way of asking search rather than a way of asking a website, and it is a different moment from a visitor speaking to a site directly.

Voice SEO

Voice SEO is the practice of optimizing content to be the answer a search engine or assistant reads aloud, and in practice it has largely merged into answer engine optimization, the broader work of being the source AI answers cite.

Zero-click search

A zero-click search is a search where the person gets their answer on the results page itself and never visits a website, so the search succeeds while the site that had the answer receives nothing.

AI disclosure and the EU AI Act

AI disclosure means telling people they are interacting with an AI rather than a human, and the EU AI Act turns that into a legal obligation in defined circumstances when its Article 50 transparency provisions become enforceable on 2 August 2026.

CCPA and CPRA

The CCPA, as amended by the CPRA, gives California residents rights to know about, delete, correct and stop the sharing of the personal information a business holds, and it asks whether that business disclosed and can service requests rather than whether it had a basis to collect at all.

Consent and call recording law

Consent in a voice deployment is three separate permissions, one to record the conversation, one to use what it produced and one to contact the person at all, and holding any of them does not grant the others.

DPDP Act

The DPDP Act is the Indian data protection law, built around consent as the principal basis for processing, which makes it narrower than GDPR in exactly the place most international deployments assume it will be looser.

GDPR

The GDPR is the European Union data protection regulation, and it sets the rules for collecting and using personal data about people in the EU no matter which country the organization holding that data is registered in.

HIPAA

HIPAA is the United States law governing the privacy and security of health information, and any voice AI vendor that handles patient information on behalf of a healthcare provider becomes a business associate directly bound by it.

PCI DSS

PCI DSS is the contractual security standard the payment card industry imposes on anyone who stores, processes or transmits cardholder data, and for a voice agent the compliant approach is almost always to make sure it never touches card details at all.

PII redaction and data residency

PII redaction removes personally identifiable information from transcripts and recordings, while data residency governs which country that data is stored and processed in, and both are standard enterprise requirements with specific consequences for a voice pipeline.

SOC 2

SOC 2 is an independent audit report, issued under the AICPA Trust Services Criteria, giving an auditor's opinion on whether the controls an organization has described are suitably designed and, in a Type II report, whether they operated effectively over a period.

TCPA

The TCPA is the United States law restricting calls and texts made with autodialers or with an artificial or prerecorded voice, and since a February 2024 FCC declaratory ruling it plainly covers AI-generated voices.

Voice biometrics and anti-spoofing

Voice biometrics identifies or verifies a person by the characteristics of their voice, and anti-spoofing is the countermeasure that tries to work out whether the voice on the line is a real person or a synthetic copy.

Watermarking of synthetic audio

Watermarking of synthetic audio embeds an imperceptible signal into generated speech so it can later be identified as machine-made, and it is the main technical answer to the question of how anyone will know what was synthesized.

Voice navigation as accessibility

Voice navigation means moving through a website by speaking instead of clicking, which gives people with motor, vision or situational limitations another way in, but it is an additional modality rather than a route to accessibility compliance: it excludes anyone who cannot speak aloud, and it has to meet accessibility standards itself.

Agent-to-agent communication

Agent-to-agent communication is AI systems interacting with each other directly rather than with a person, and in voice the near-term version is a customer's assistant dealing with a business's assistant.

Ambient agents

An ambient agent works continuously in the background rather than waiting to be asked: it observes a stream of events, decides when something is worth acting on, and surfaces only when it judges itself useful.

Emotion detection and affective computing

Emotion detection attempts to infer a person's emotional state from signals such as their voice, and affective computing is the broader field of systems that recognize, interpret or simulate emotion; the acoustic measurement is real, but the inference from it to a named emotion is scientifically contested.

Model Context Protocol

The Model Context Protocol (MCP) is an open standard, introduced by Anthropic in November 2024, that connects AI systems to external tools and data through one consistent interface instead of a bespoke integration for every pairing.

Sentiment analysis

Sentiment analysis determines whether what someone said was positive, negative or neutral by classifying the language they used, not by trying to read the speaker's emotional state.

Voice-first browsing and AI browsers

Voice-first browsing means using a website by speaking instead of clicking, and AI browsers are browsers with a built-in assistant that reads, summarizes and increasingly acts on pages for the visitor.

Pricing models for voice agents

Pricing models for voice agents are the ways a vendor charges, by the minute, by the conversation, by the outcome or by seat, and the model chosen shapes vendor and customer incentives more than the rate does.

Unit economics of voice agents

Unit economics of a voice agent is the delivered cost of a single conversation weighed against what that conversation earns, and it carries more weight here than in most software because each exchange runs up a genuine variable cost that scale never removes.

Voice AI market sizing

Voice AI market sizing is the practice of estimating how large the voice AI category is and how fast it is growing, where the wide gap between research firms is the most useful thing the figures tell you.