Electric Twin, a technology company developing AI-driven synthetic audiences to predict real-world behaviour, announced $14 million in total funding. The round was led by Atomico with participation from LocalGlobe, Mercuri and Marc Andreessen, following a $4 million pre-seed.
Building Synthetic Audiences to Predict Human Behaviour
Founded after witnessing high-stakes pandemic decisions made without reliable tools to anticipate public response, Electric Twin builds simulations powered by large language models, social science and machine learning. As Co-Founder Ben said: “a pile of facts is no more science than a pile of stones is a house.” Professor Michael Muthukrishna of the London School of Economics serves as scientific advisor.
Academic research with LSE found the platform produced insights 10,000 times faster than traditional market research at 95% accuracy across nearly 40,000 evaluations in 155 countries. In a commercial benchmark of more than 8,000 respondents, predictions were indistinguishable from conventional methods, with a mean absolute error of 0.05.
Funding to Scale Hiring, Transparency and Market Expansion
The company said accuracy is critical. Without it, synthetic audiences are just a faster way to be wrong.
Funding will expand hiring, scale go-to-market and deepen evaluation and transparency. Electric Twin frames the mission as closing the understanding gap between decision-makers and the people affected by decisions, captured in the line: Ask them anything.