Element Human
Building Element Human's AI-Powered Emotion Detection MVP
Vodworks helped Element Human (UK) build an AI-powered Facial Expression Recognition (FER) MVP in just 12 weeks, capturing real-time emotional responses across six emotions. The MVP became the foundation of Element Human's Engage Platform, enabling emotion-driven insights for businesses across media, advertising, gaming, and UX testing.
Challenge
Element Human needed to rapidly build a working FER MVP on a tight timeline without extensive internal hiring or delays. The MVP needed to capture real-time emotional responses, process data immediately for ML model training, and be optimised for mobile performance — all while maintaining high detection accuracy across diverse datasets.
Solution
Vodworks assembled a dedicated team of developers and AI specialists with FER expertise. They optimised emotion detection accuracy and mobile performance, integrated BBC-provided ground data to improve model training across diverse datasets, and built real-time data processing capabilities to generate actionable insights. The MVP captured real-time emotional responses across six emotions (happiness, sadness, surprise, fear, disgust, and anger) to serve as the foundation for the Engage Platform.
Results
MVP delivered in 12 weeks, enabling rapid testing and iteration. The MVP validated FER real-time capabilities and demonstrated its potential to enterprise clients. The success directly led to Element Human's full Engage Platform, now used across media, advertising, gaming, and UX testing industries. Element Human was able to move forward with confidence without extensive internal hiring.
Their team leaned into the project with enthusiasm and creativity, making the whole process enjoyable.
Matt Celuszak, CEO, Element Human
