Airwordo

Airwordo was an AI-powered Android application designed to automatically generate live captions from any audio or video playing on a user’s device. The product targeted advanced language learners and accessibility use cases, turning spoken content into real-time text to support vocabulary building and content comprehension.

Our team built AirWordo as an in-house startup, going through the full product lifecycle from idea validation to technical implementation. We leveraged early-stage speech recognition technology, Android system-level accessibility features to capture audio from multiple sources, and on-device audio processing combined with AI transcription models. Beyond development, we conducted user interviews, validation experiments, accelerator programs, and product pivots, ultimately making a data-driven decision to close the project due to limited market potential. This experience gives us first-hand understanding of startup risks, MVP trade-offs, and technical experimentation, making us stronger partners for founders navigating early-stage product decisions.