Reading up on Deepgram

Anthropic is building a Mac‑first meeting recorder called Parka inside Claude Desktop that captures audio and turns meeting action items into tasks for Claude Cowork and Claude Code agents.

James Wang provides step-by-step instructions for non-technical users to create AI agents using Claude and ChatGPT Projects, covering language learning, morning briefings, and meeting transcription pipelines.

By piping Deepgram Flux, Groq's LLM, and ElevenLabs TTS with geographic optimization, the author achieved ~400ms end-to-end voice agent latency, beating Vapi by 2x.

In 2025, voice AI APIs cost $5–$32/hour, making them more expensive than human call center workers in most developing countries until inference costs drop around 2030.