Skip to main content
AI glossary

Understand the concepts behind AI agents and realtime communications.

Use the MobDial AI glossary as an entry point to the terms behind conversational agents, speech systems, models, tools, retrieval, evaluation, routing, guardrails, and human handoff.

One MobDial identityRole-aware workspacesAPIs + automation
MobBot presenting MobDial AI agent orchestration

AI Agent Studio

Product workspace preview

Test mode

Agent

Configured

Tools

Connected

Handoff

Ready

Live workflow

ListenIntent and customer context resolved
ReasonInstructions, knowledge, and tools evaluated
ActRespond, use a tool, or hand off with context

Execution guardrails

  • Approved knowledge only
  • Scoped tool permissions
  • Test before publish
  • Human escalation path

Agents

Understand instructions, tools, memory, context, handoff, and agent lifecycle concepts

Speech

Learn TTS, STT, latency, turn-taking, voice resources, and transcript terminology

AI systems

Learn models, inference, embeddings, retrieval, evaluation, and observability concepts

Safety

Understand permissions, consent, provenance, approvals, and human oversight terminology

Core concepts

A shared vocabulary for builders, operators, and business teams.

The glossary groups terminology around the systems people actually configure and operate inside MobDial.

Agents

Agent orchestration

Instructions, context, tools, memory, state, planning, actions, escalation, and handoff.

Explore
Speech

Speech systems

Text to speech, speech to text, realtime audio, voice resources, turn-taking, and transcripts.

Explore
Models

Models and inference

Model providers, inference, tokens, context windows, embeddings, routing, and structured outputs.

Knowledge

Retrieval and grounding

Knowledge sources, retrieval, embeddings, citations, freshness, and approved business context.

Quality

Evaluation and observability

Test cases, traces, latency, quality metrics, failure analysis, and release evidence.

Safety

Guardrails and approvals

Permissions, consent, policy checks, provenance, high-risk actions, and human review.

Explore
Learning path

Learn the term. See it in context. Apply it safely.

01

Start with the concept

Understand the term and the problem it describes.

02

Connect it to the product surface

See where the concept appears in agents, speech, developer, workflow, or operations experiences.

03

Verify implementation details

Use current documentation and your configured environment for exact API, provider, and deployment behavior.

Next step

Turn AI terminology into practical MobDial workflows.

Use the glossary with product documentation, developer references, and safety guidance when designing or operating AI-enabled communications.