OriginalVoices provides an AI-powered audience research platform that lets companies query Digital Twins of real people—40,000+ individuals who have trained and own their AI representations—to get real-time human perspectives on products, messaging, and ideas. The platform integrates via API, SDK, or MCP connectors into tools like Claude, enabling teams to describe any audience in plain language and receive structured answers in under 20 seconds. Every response is grounded in a real person who remains in the loop to validate and update their Twin.
Funding
£1M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
ISFounders
Product
Problem
Traditional audience research relies on web scraping, which only surfaces existing online content, or synthetic personas built from aggregated surveys and outdated data—neither of which reflects what a specific audience actually thinks or feels. Without a fast way to hear from real people, teams fall back on assumptions and guesswork, shipping products and messaging based on intuition rather than verified human perspective.
Solution
OriginalVoices gives AI agents and assistants access to real-time human insights by querying Digital Twins—AI representations trained, owned, and validated by the 40,000+ real people they represent. Users describe their target audience in plain language and ask any question, receiving structured responses in seconds via API, SDK, CLI, or MCP integrations that work with tools like Claude. The platform eliminates the need for panel recruitment or audience setup, letting teams test concepts, explore pain points, and validate messaging directly against real human perspectives. Every answer is grounded in a real person who remains in the loop to train, update, and validate their Twin, ensuring insights are current and authentic rather than simulated or statistically guessed.
Target Audience
Primary customers are product teams, marketers, and AI developers who need fast, real human feedback to ground AI outputs, validate messaging, and shape product decisions—particularly those building AI agents or assistants that require up-to-date audience insights.
Features
- Digital Twin network of 40,000+ real people, each owning and training their AI representation for authentic, verifiable responses
- Natural-language audience definition—describe any target segment without panel recruitment or setup
- Sub-20-second average response time for real-time research queries
- REST API, CLI, Vercel AI SDK tool, and self-hosted MCP server for integration into custom AI applications and agent frameworks
- Native MCP connector for Claude, enabling in-conversation primary research for content creation, positioning, and market analysis
- Structured JSON responses designed for AI agents to consume and act upon directly
- 320k+ individual Twin answers available for querying across diverse topics and audiences