Remesh launches AI-moderated video interviews with Remy

Remesh on October 20, 2026 launched Remy Interviews, a new way for research teams to run one-on-one video interviews moderated by its AI research agent. The product is designed to help teams scale in-depth qualitative research without giving up human oversight.

Why it matters: - Remy Interviews targets a core bottleneck in research: in-depth interviews are valuable, but they are time-consuming to staff and scale. - The new format lets teams run more conversations from one discussion guide while keeping a human researcher in control. - Remesh is positioning the product as a way to preserve interview depth without adding the usual scheduling and moderation burden.

What happened: - Remesh announced Remy Interviews, a new capability inside its conversational research platform. - The tool runs one-on-one video interviews moderated by Remy, Remesh’s AI research agent. - Remy asks questions aloud, listens to spoken answers, and generates follow-up questions in real time. - Teams can schedule up to 10 interviews in a single conversation using the same discussion guide. - The launch was announced on October 20, 2026.

The details: - Researchers can choose either a human moderator or Remy AI for each interview. - Remy AI sessions can be created without a firm date and time, letting participants complete interviews on their own schedule. - Researchers can write a discussion guide or build one with Remy. - The Flex option lets participants join at a day and time of their choice within a one-year window from study publishing, or by a custom end date. - Probing depth can be set to Off, Standard or Deep. - Focused follow-ups can steer Remy toward examples and stories, barriers or comparisons, or a custom focus area. - The focus area is not shown to participants. - Teams can choose from three voices for Remy and preview each one before publishing. - An optional anonymize setting makes sessions audio-only and assigns participants a generic name for sensitive topics. - Simulate mode launches a 15-minute practice interview immediately and supports testing with up to 10 collaborators and 10 of the researcher’s own participants before going live. - Participants join through a browser link with no download required. - Participants are told before the session that Remy is an AI interviewer, that the session is recorded and transcribed, that a human researcher may join or take over, and that they can leave at any time. - Closed captions are available, and participants can ask Remy to repeat any question. - Remy’s follow-up questions are based on what participants actually say. - In one sustainability study example, a participant who mentioned beach cleanups and recycled materials was asked what made those claims feel trustworthy rather than like marketing. - Researchers and stakeholders can join live, see Remy’s progress through the guide and the elapsed time, pause Remy, or take over the session. - Existing Remesh tools, including a private backroom, notes and moment marking, remain available during sessions. - Recordings and transcripts feed into Remesh’s video analysis tools for AI-generated summaries, moments and themes, custom themes and highlight reels.

Between the lines: - Remesh is aiming to remove the moderator bottleneck without removing human judgment from the process. - The product blends automation with oversight, which could appeal to teams that want more interviews without fully delegating qualitative research to AI. - The explicit participant disclosure and takeover options suggest the company is trying to make the experience feel transparent rather than fully autonomous.

What's next: - Remesh is inviting prospective customers to schedule a demo to learn more about the offering. - The company says researchers can step in at any moment, signaling that live human moderation will remain part of the workflow for teams that need it.

The bottom line: - Remy Interviews gives research teams a faster way to run repeated one-on-one video interviews while keeping the researcher in the loop.