Customer Interview API for AI Agents: How Agents Run Real Research

An AI agent that can ask real customers questions is categorically more useful than one that can only reason about them. Getting there requires an interview API that supplies participants, runs moderated research-grade conversations, and returns structured findings — not just a voice synthesis layer the agent talks into.

This guide covers how the customer interview API pattern works, the three integration patterns agents commonly use, how recruitment operates programmatically, and how User Intuition’s MCP server differs from voice AI APIs like ElevenLabs and Vapi. For full API documentation including tool schemas, see docs.userintuition.ai/mcp-server/overview.

What a customer interview API does, vs what survey APIs do

A survey API delivers questionnaires — structured prompts with fixed-choice or short-text responses. Survey APIs are fast, cheap per response, and statistically tractable at large n. They are the right tool when you need to measure: what percentage of users would use a feature, how satisfaction scores track over time, whether a respondent belongs to a behavioral segment.

A customer interview API does something different. It launches AI-moderated conversations that probe, follow up on unexpected answers, and surface the reasoning and mental models behind behavior. The output is not a distribution of choices — it is a corpus of explained positions that reveals why users behave the way they do and what they need from a solution that does not yet exist.

The tradeoff is cost and volume. Survey APIs work for large-n quantitative signals at low cost per response. Interview APIs work for understanding at the depth that changes product and positioning decisions. Both are part of a complete research stack; they are not substitutes.

For AI agents specifically, the relevant capability gap is this: a survey API cannot probe. When a respondent says “I prefer option B,” a survey records that. A moderated interview follows up: “What about option A didn’t land?” “What would option B need to do for it to be worth the switch?” The AI moderator extracts the reasoning, not just the choice. That reasoning is what changes decisions.

Three integration patterns

A customer interview API supports three distinct integration patterns, each matched to a different decision velocity and research depth requirement.

Pattern 1: Human Signal — rapid panel polls

ask_humans dispatches a question to a panel segment and returns structured responses. No study configuration, no discussion guide, no recruitment pipeline — a single tool call and answers arrive within hours.

When to use it: Directional checks where a full moderated study is more rigor than the decision requires. Headline preference. Claim credibility. Simple concept reaction. The agent has a question that needs real signal, not an LLM simulation of signal, but does not need to understand why.

Example agent sequence:

ask_humans(
  question="Which headline do you prefer for a B2B SaaS product?",
  options=["Ship 10x faster" , "Build with confidence"],
  segment="product_managers_saas",
  sample_size=20
)
→ get_results(study_id=...)
→ Returns: preference split + representative explanations

Results within hours. Cost: pay-per-use credits at $12.50 per chat response.

Pattern 2: Full AI-moderated study

create_assistant configures an AI moderator with a discussion guide, study type, and language. bulk_create_invites_from_segment recruits qualified participants from the panel. The AI moderator conducts the interviews. analyze_transcripts returns structured findings when fielding completes.

When to use it: Research questions where understanding why is required — churn reasons, concept validation, competitive positioning, UX friction mapping. The agent needs more than a preference split; it needs the reasoning, the objections, and the unarticulated needs.

Example agent sequence:

create_assistant(
  study_type="concept_test",
  discussion_guide="...",
  language="en",
  mode="audio"
)
→ bulk_create_invites_from_segment(
    assistant_id=...,
    segment_filters={role: "product_manager", company_size: "100-500"},
    count=15
  )
→ [interviews run, AI moderated]
→ analyze_transcripts(assistant_id=...)
→ Returns: theme summaries + verbatims + segment findings

Results in 2-3 hours for 15 participants. Cost: $20 per audio interview on Pro plan.

Pattern 3: Intelligence Hub queries

query_intelligence retrieves synthesized answers from accumulated research history. No new study — the agent asks a question over everything the workspace has already learned.

When to use it: The research question has probably been answered before, or the agent needs context from prior studies to design the next one. Intelligence Hub queries run in seconds and cost nothing beyond the prior studies that populated the hub.

Example:

query_intelligence(
  query="What objections come up most often when SMB buyers evaluate pricing?"
)
→ Returns: synthesized findings with source references from prior studies

Pattern 3 is the compounding asset. Every study run via Patterns 1 or 2 adds to the hub. An agent with 50 studies behind it answers Pattern 3 queries with far higher confidence than one with 5. The hub is what converts a series of one-off research projects into an institutional knowledge base that agents can draw on indefinitely.

Recruitment via API

The participant supply is what separates a research API from a voice AI API. bulk_create_invites_from_segment accepts targeting parameters and generates invitations for qualifying participants from the User Intuition 4M+ panel.

Available targeting dimensions:

• Job role and seniority
• Company size and industry
• B2B segment (company attributes) vs B2C segment (demographic + behavioral)
• Geography and language (50+ languages supported)
• Custom screening criteria via create_panel_survey

For bring-your-own-participant workflows — customer advisory panels, existing CRM contacts, community members — create_invite handles individual invitations with custom contact details. list_external_participants and trigger_integration_sync support CRM-sourced recruitment from Salesforce and HubSpot.

send_reward handles incentive distribution once interviews complete. Reward amounts are configured at the study level and distributed automatically to participants without any manual step.

How does User Intuition handle agent-driven customer interviews?

User Intuition provides the production-grade customer interview API for AI agents — 72 MCP tools covering every phase from rapid poll to cross-study synthesis. The tool surface maps directly to the three patterns above: Human Signal group (ask_humans, get_results) for Pattern 1; Studies, Invites and Participants, and Calls and Interviews groups (create_assistant, bulk_create_invites_from_segment, analyze_transcripts) for Pattern 2; Intelligence Hub group (query_intelligence, generate_report, generate_powerpoint) for Pattern 3.

The 4M+ panel covers B2B and B2C across 50+ languages, with full targeting by role, company size, industry, and behavioral attributes. Studies start at $200, return results in 24-48 hours for standard runs, and carry 5/5 ratings on G2 and Capterra. The MCP server is available via stdio (install: npx -y @userintuition-ai/mcp) for local agent development and via Streamable HTTP at https://mcp.userintuition.ai/mcp for production deployments — both use a single ui_sk_ API key with no scope configuration. The Starter plan is free with three interviews on signup; no credit card required to test the integration.

Comparison: voice AI APIs vs research-specific APIs

Voice AI platforms — ElevenLabs, Vapi, and similar — provide real-time voice synthesis and conversation infrastructure. They are the right choice when the product is the conversation: customer support agents, voice assistants, interactive voice response systems. The API surface is optimized for real-time call handling: low-latency audio generation, call routing, transcript capture.

A research-specific API is built for a different use case. The conversation is a means to an end — the end being structured qualitative evidence. The distinctions that matter for research:

For agents that need to talk to customers the product already has, a voice AI API can be appropriate. For agents that need to recruit qualified strangers, conduct research-methodology conversations, and return analyzable evidence — the research-specific API is the correct layer.

Neither replaces the other. Some agent architectures use a voice AI layer for real-time customer support and a research API for periodic research programs. The distinction is: do you need participants, or do you already have them? Do you need research methodology, or is any conversation sufficient? Do you need structured findings, or is raw transcript enough?

Get an API key

For MCP integration (recommended for AI agents):

1. Sign up at app.userintuition.ai/sign-up — Starter plan, free, three interviews on signup, no card.
2. Generate a ui_sk_ key from Settings > API Keys.
3. Connect via stdio (npx -y @userintuition-ai/mcp) or Streamable HTTP (https://mcp.userintuition.ai/mcp).
4. Full tool reference: docs.userintuition.ai/mcp-server/overview.

For a broader overview of the agentic research methodology — what agents can accomplish once connected and how the Intelligence Hub compounds over time — see agentic research.