Glossary
Definitions of the terms used throughout the MiaRec documentation.
Conversations and their data
| Term |
Definition |
| Conversation |
The unit of capture and analysis in MiaRec: a recorded interaction between your organization and another party — today, a recorded phone call. |
| Recording |
The captured audio (or screen) file of a conversation. Used when the media file itself is meant; the interaction as a whole is a conversation. |
| Screen recording |
A capture of the agent's desktop screen during a conversation, linked to that conversation. |
| Grouped conversation segments |
Related recordings (from transfers, holds, and resumes) linked together and presented as parts of one conversation. |
| Transcript |
The speaker-separated, timestamped text of a conversation, produced automatically by speech-to-text. |
| Summary |
A short AI-written digest of a conversation: reason, key points, outcome, next steps. |
| Tag |
A label users apply to conversations to categorize and filter them. Tags can be organized hierarchically. |
| Note |
A free-text comment attached to a conversation. Notes support replies and can be anchored to a specific moment in the transcript. |
| Custom field |
A data field your organization adds to conversations, filled manually, by an integration, or by an AI task. |
| Client |
A directory record for a customer or account your organization talks to — a name plus phone numbers. Conversations are matched to clients by phone number. |
| Confidential |
A flag restricting a conversation's visibility to users with confidential-conversation access. |
| Do not delete |
A flag protecting a conversation from automatic retention cleanup. |
AI analytics
| Term |
Definition |
| Conversation Analytics |
The MiaRec product that automatically analyzes conversations: transcription, summaries, sentiment, topics, insights, and CX metrics. |
| Insight |
A piece of information the AI extracts from a conversation, such as the call reason or a satisfaction score. |
| AI task |
A configured instruction (prompt) that produces an insight and stores the result — as a summary, sentiment score, topic match, custom field value, or Auto QA evaluation. |
| Sentiment score |
An AI-assessed score of a conversation's emotional tone, from -100 (most negative) to +100 (most positive), assessed separately for the conversation overall, the agent, and the customer. |
| Topic |
A named subject your organization wants to detect in conversations. The AI identifies which topics each conversation discusses. |
| CX metrics |
Customer-experience scores — such as CSAT, NPS, or NES — estimated per conversation by AI tasks your organization configures. |
| Entity recognition |
Detection of specific information types in transcripts (names, dates, amounts, card numbers, and similar). Abbreviated NER. |
| Redaction |
Removal of sensitive content from conversations: masked in the transcript and silenced in the audio, according to rules your administrator defines. |
Quality Assurance
| Term |
Definition |
| Quality Assurance (QA) |
The MiaRec product for evaluating conversations against your quality standards — Auto QA first, manual evaluation second. |
| Scorecard |
The evaluation template: weighted sections and questions with a passing threshold. (Shown as evaluation form in some older parts of the product.) |
| Auto QA |
AI-powered scoring of conversations against a scorecard, with a justification and evidence for every answer. |
| Manual evaluation |
An evaluation completed by a person, question by question. |
| Evaluation |
A scored assessment of a conversation, produced by Auto QA or by an evaluator. |
| Override |
A reviewer's correction of an AI answer on an auto-scored evaluation. The original AI answer is preserved alongside the override. |
| Passing threshold |
The minimum score, set on the scorecard, that an evaluation must reach to pass. |
People and access
| Term |
Definition |
| Your organization |
Your company's own environment in MiaRec — its users, conversations, and settings. |
| User |
Any person with a MiaRec login. |
| Customer |
An end customer of your organization — the other party in a conversation. Customers are not MiaRec users. |
| Agent |
A user whose conversations are captured and analyzed. |
| Supervisor |
A user who manages a team: reviews conversations, evaluations, and dashboards for their groups. |
| Administrator |
A user who configures your organization: accounts, roles, AI settings, and data management. |
| Group |
A set of users within your organization, usually a team; used for dashboards, access scoping, and license tracking. |
| Role |
A named set of permissions assigned to a user. |
| Permission |
An individual capability a role grants, such as playing audio or evaluating conversations. |
| Access scope |
The data-visibility level of a role: Organization, Selected Groups, or User. |
| Term |
Definition |
| Ingestion (or capture) |
How conversations enter MiaRec: real-time recording, telephony platform sync, bulk import, or manual upload. |
| Live monitoring |
Listening to a user's active conversation in real time; requires a monitoring license. |
| Extension |
A telephony endpoint identifier (extension or phone number) assigned to a user so their conversations are matched to them. |
| Retention |
Automatic deletion policies applied to aged conversations. The do not delete flag exempts a conversation from retention. |
| License (or seat) |
A per-user entitlement to a capability such as recording, conversation analytics, quality assurance, live monitoring, or screen recording. |
| Subscription |
The set of MiaRec products and capacity your organization has purchased. See Subscriptions & Licensing. |
| Service provider |
The party operating the MiaRec platform for your organization, when MiaRec is delivered as a service. Some platform settings are provider-managed. |
| MCP (Model Context Protocol) |
A standard interface through which external AI tools can query your MiaRec data. |