We can certainly handle the integration, but it will be time-consuming due to technical complexities. As an orchestration platform, we manage interactions between different components, each with its own category. For example, OpenAI, Claude, and other LLM providers fall under the “LLM” category. Deepgram, Gladia, and Whisper are in the “Transcriber” category, while 11labs, PlayHT, and others fit into the “Voice Provider” category. However, Hume is different. It doesn’t strictly fall into the “Voice Provider” section since it doesn’t offer standalone TTS. Instead, they package multiple services, including LLM and voice functionalities, making it more than just a typical voice provider. Integrating Hume means creating an entirely new component for it. We’re also waiting for the release of a speech-to-speech model from OpenAI or another provider, which would help streamline this process and resolve many of the challenges we’re currently facing.