Terapage AI-Moderated Telephone Interview

Terapage AI-Moderated Telephone Interviewing, also known as Terapage Voice Research or AI CATI, allows research teams to conduct telephone interviews at scale using AI-assisted voice moderation. Researchers can upload participant lists, configure the AI interviewer, define the interview script, set call rules, schedule outbound calls, capture consent, record interviews, transcribe conversations and generate insights from completed calls. It is designed to support modern telephone research without relying entirely on manual call-centre operations.

It serves a similar research purpose, but it is more automated and AI-assisted. Traditional CATI usually relies on human interviewers calling participants and manually following a script. Terapage’s AI-moderated telephone interview feature allows the AI interviewer to place calls, introduce the study, capture consent, ask questions, probe responses, manage call flow and produce transcripts and analysis. Human researchers can still remain in control of campaign setup, scripts, compliance rules, review, reporting and, where enabled, live handover.

No. The feature is designed to assist research teams, not replace them. Researchers still design the study, define objectives, upload or manage participants, create interview scripts, configure consent, set call rules, review outputs and interpret findings. The AI helps with scalable interview execution, transcription, structured probing, call handling and analysis support.

A researcher creates a telephone interview campaign, uploads participants, selects or provisions a phone number, configures the AI interviewer, adds the interview script, sets consent wording, defines call windows and retry rules, then launches the campaign. Terapage queues the calls, places outbound calls through the connected telephony provider, captures consent, conducts the AI-moderated interview, records and transcribes the call where enabled, and then generates summaries, themes, sentiment and other research outputs.

Yes. Researchers can select or buy supported phone numbers for telephone interview campaigns. The system can support number search by country, region, city or area code, number assignment to a campaign, caller ID branding where available, and number release after the campaign ends. Some countries may require additional verification, address details or business documentation before numbers can be activated.

Yes, number porting can be supported for enterprise or eligible clients. Porting usually requires documentation such as a letter of authorisation, a recent phone bill and business verification documents where applicable. Porting is not instant and depends on carrier approval, local requirements and number portability rules in the relevant country.

Yes. Researchers can upload participant lists using formats such as CSV or Excel. Typical fields may include first name, last name, phone number, country, timezone, preferred language, segment, consent status, custom attributes, external ID and notes. The system should validate phone number format, country code, duplicate records, missing required fields, invalid numbers, unsupported countries and do-not-call matches before launch.

Yes. The participant upload process can check for missing numbers, duplicate phone numbers, invalid formats, unsupported countries and do-not-call conflicts. This helps reduce wasted call attempts, unnecessary costs and poor campaign performance.

Yes. Researchers can enable a pre-call message by SMS, WhatsApp or email depending on configured channels. This can inform participants that they may receive a research call, explain the purpose of the call and provide opt-out instructions. Pre-call messages can help improve answer rates, reduce confusion and support better consent transparency.

Yes. Researchers can configure the AI interviewer’s voice, language, accent, tone and personality. For example, the interviewer may be set to sound warm and professional, calm and clinical, neutral and academic, friendly and conversational, or more formal for compliance-sensitive studies. This allows the interview experience to match the study type, audience and brand tone.

The feature is designed to support multiple languages and accents depending on the available AI voice stack. Examples may include British English, American English, Indian English, Nigerian English, French, Spanish, Arabic and other supported languages. Language availability may vary depending on voice provider capability, transcription quality and campaign requirements.

Yes. Researchers can build telephone interview scripts with an introduction, consent statement, screening questions, main questions, probing instructions, skip logic, termination logic, thank-you message, incentive message, opt-out instructions and human handover triggers. The AI can use the script to conduct a structured interview while still probing naturally where appropriate.

Yes. The AI can use adaptive probing rules to ask relevant follow-up questions based on participant responses. For example, it can ask for clarification, request an example, probe the reason behind an answer, or explore a topic in more detail. This helps make the interview feel more natural than a rigid telephone survey.

Yes, where barge-in support is enabled. If a participant starts speaking while the AI is talking, the system should detect the interruption, stop the AI voice playback and listen to the participant before continuing with a context-aware response. This is important for making telephone interviews feel more natural and less robotic.

Before the substantive interview begins, the AI can read a consent statement and ask the participant to confirm whether they agree to continue. Participants may be able to give consent verbally or by pressing a keypad option, depending on the configuration. If consent is not given, the call should end politely, the status should be recorded as no consent, and the interview should not proceed.

Yes, calls can be recorded where enabled and legally appropriate. If recording is switched on, the participant should be informed before the substantive interview begins. Recording settings should be controlled by the researcher or account administrator and supported by audit logs. Recording laws may vary by country, so researchers are responsible for ensuring lawful consent and compliance.

Yes. Participants can opt out during the call by saying phrases such as “stop,” “remove me,” “do not call me again” or “opt out.” The system will stop the interview, add the participant to a do-not-call list, prevent future calls and log the opt-out event.

Yes. Do-not-call lists can be maintained at workspace, campaign, account or phone-number level. Before calls are placed, the system checks uploaded participants against relevant DNC lists to avoid contacting restricted numbers.

Yes. Researchers can configure call windows by date, time, timezone and day of week. They can also define quiet hours, retry delays, maximum call attempts, no-call days and concurrency limits. This ensures participants are contacted at appropriate times and improves response rates.

Retry rules can be configured by the researcher or administrator. A typical setup may allow up to three attempts with delays between retries. The system stops retrying after completion, opt-out, refusal, invalid number or final call status.

If a participant does not answer the call, the system marks the attempt as “no answer”. Based on campaign rules, the system may retry the participant after a defined delay. Researchers can configure how many retries are allowed and the time gap between attempts to optimize reach and reduce unnecessary call volume.

If the system detects voicemail, it avoids starting the full interview. Depending on campaign configuration, it may either leave a short pre-recorded message or simply mark the attempt as voicemail and schedule a retry. This helps preserve interview quality and avoids collecting incomplete responses.

Yes. Inbound callback handling is supported. When a participant calls back, the system can identify the campaign via the assigned number and match the caller to a participant record if available. It may resume the interview, offer rescheduling, capture opt-out requests, or route the call based on campaign rules.

If the caller is not matched to any participant record, the system follows predefined handling rules. This may include playing an informational message, capturing voicemail-style responses, logging the call for review, or routing it to a fallback flow depending on campaign configuration.

Yes. Human handover is supported where enabled. A researcher can join a live call if the participant requests human assistance, if the AI confidence is low, if escalation rules are triggered, or if manual intervention is required. The call remains recorded and logged for audit and research integrity.

After each completed call, the system can generate transcripts, summaries, key themes, sentiment analysis, emotional indicators, key quotes, pain points, user needs, objections and structured research insights. These outputs help researchers quickly move from raw conversations to actionable findings.

Yes. Researchers can access call recordings, transcripts, summaries and analysis outputs through the campaign dashboard. Access is controlled by permissions, and data retention policies may apply depending on workspace configuration.

Campaign dashboards can display key performance indicators such as total participants, queued calls, completed calls, answered calls, no answers, failures, opt-outs, consent declines, completion rate, answer rate, average call duration, retry success rate and estimated cost metrics. These insights help researchers monitor performance in real time.

Yes. The system tracks costs including phone number rental, call duration, recording usage, transcription processing, AI analysis and human handover time where applicable. Both estimated and actual costs can be reviewed at campaign level.

Yes. Researchers can set budget caps and usage limits to prevent overspending. This is especially useful for large-scale campaigns with high retry volumes or multiple concurrent calls. When limits are reached, the system can pause or restrict further calling activity.

Terapage uses multiple safeguards to protect number reputation, including controlled call pacing, retry limits, number rotation, number pools, opt-out handling, and monitoring of call behaviour patterns. Where available, branded calling and verified caller identity systems may also be used.

Yes, where supported by the telecom provider and region. Caller ID branding may allow participants to see a verified business name or organisation identity instead of an unknown number. Availability depends on country regulations, carrier support and number type.

Yes. The system is designed for scalable telephone interviewing. It supports concurrent call limits, queue management, number pools, scheduling rules and performance controls to manage high-volume research campaigns efficiently.

The system uses programmable telephony for outbound and inbound calling, real-time audio streaming, speech-to-text transcription, AI voice generation, an interview orchestration engine, secure storage systems and analytics infrastructure to process and analyse interview data.

The supported coverage list includes Albania, Angola, Argentina, Barbados, Belize, Bermuda, Bolivia, Bosnia and Herzegovina, Colombia, Dominican Republic, Finland, French Guiana, Guadeloupe, Iceland, Israel, Jordan, Lebanon, Lithuania, Martinique, Mauritius, Mayotte, Mexico, Moldova, Monaco, Morocco, North Macedonia, Panama, Paraguay, Réunion Island, Saint Barthélemy, Saint Martin, Serbia, Slovenia, Ukraine, Zimbabwe, United Kingdom, United States, US Virgin Islands, Canada, Puerto Rico, Denmark, New Zealand, Costa Rica, Chile and Cyprus.

Countries currently supporting all number types include Angola, United States, US Virgin Islands, Canada and Puerto Rico depending on availability and carrier rules.

Local numbers are available in countries including Albania, Argentina, Barbados, Belize, Bermuda, Bolivia, Colombia, Dominican Republic, Finland, French Guiana, Guadeloupe, Iceland, Israel, Jordan, Lebanon, Lithuania, Martinique, Mauritius, Mayotte, Mexico, Moldova, Monaco, Morocco, North Macedonia, Panama, Paraguay, Réunion Island, Saint Barthélemy, Saint Martin, Serbia, Zimbabwe, United Kingdom, New Zealand, Costa Rica, Chile and Cyprus.

National numbers are available in Argentina, Barbados, Belize, Bermuda, Bolivia, Bosnia and Herzegovina, Colombia, Dominican Republic, French Guiana, Guadeloupe, Iceland, Israel, Jordan, Lebanon, Martinique, Mauritius, Mayotte, Moldova, Monaco, Morocco, North Macedonia, Panama, Paraguay, Réunion Island, Saint Barthélemy, Saint Martin, Serbia, United Kingdom, Costa Rica and Cyprus.

Mobile numbers are currently supported in Colombia, Ukraine, United Kingdom and Denmark. Some countries offering “all types” may also support mobile numbers depending on availability.

Toll-free numbers are supported in Albania, Argentina, Bermuda, Bosnia and Herzegovina, Colombia, Iceland, Mexico, Moldova, Monaco, Morocco, North Macedonia, Réunion Island, Saint Barthélemy, Saint Martin, Serbia, Slovenia, Ukraine, Denmark, New Zealand, Chile and Cyprus.

No. Finland currently supports local numbers only and does not support mobile number provisioning.

Slovenia currently supports toll-free numbers only.

Zimbabwe currently supports local numbers only.

Yes. Number availability can change depending on carrier inventory, regulatory requirements, country rules, address verification needs and telecom provider updates. Live availability should always be confirmed during provisioning.

No. Some countries allow instant provisioning, while others require manual approval, documentation, or address verification. Availability also depends on number type and carrier inventory.

Yes. Some countries require business registration documents, proof of address, authorisation letters, identity verification or telecom compliance paperwork before numbers can be activated.

Yes. Campaigns can use number pools across multiple countries where supported, depending on configuration and compliance rules. This helps improve reach across regions and time zones.

Yes. Calls can be scheduled and routed based on participant timezone, country or region to ensure calls occur at appropriate local times.

Yes, but the system will flag unsupported countries before launch. Researchers may need to remove or adjust such entries before proceeding.

Terapage provides compliance controls such as consent capture, opt-out handling, DNC lists, call windows, audit logs, recording controls and data retention policies. However, researchers remain responsible for ensuring compliance with local laws and regulations.

Yes. Results can be exported as CSV, Excel, transcripts, audio recordings, PDF reports, theme analysis, participant datasets and cost summaries depending on system configuration.

No. It can also be used for customer experience research, employee feedback, policy studies, academic research, product testing and other structured telephone-based insights.

It combines AI interviewing, telephony, transcription, analytics, participant management, consent capture, compliance tools and reporting into a single integrated workflow for scalable research.

Upload your participants, choose your interviewer voice, set the call rules, and let Terapage conduct, transcribe, analyse and report telephone interviews at scale.