Kapwing, a collaborative video editor designed to help teams edit videos faster, has successfully built AI-first features by partnering with AssemblyAI, an AI company that provides automatic speech-to-text and other AI-powered tools. Kapwing's CTO, Joshua Grossberg, emphasizes the importance of considering user needs when building AI features, such as word-by-word timestamps and translation, to improve workflow efficiency and revenue growth. Kapwing has partnered with AssemblyAI to address accuracy issues with their previous API and is now utilizing AssemblyAI's Core Transcription AI model to enhance transcription editing capabilities, including precise word timings and foreign language translations. The company plans to leverage AI in future features, such as automatically generating highlights or teasers, to simplify the video creation process for businesses and creators.