Show HN: Open-source customizable AI voice dictation built on Pipecat (github.com)

The open-source project Tambourine has been launched, offering a customizable AI voice dictation tool built on the Pipecat framework. This innovative application serves as a universal voice interface, allowing users to dictate text into any software, including email clients, code editors, and terminals, by simply speaking. Aimed at enhancing productivity, Tambourine allows users to speak at an average speed of 130-160 words per minute compared to the 40-50 words per minute of typing, effectively capturing ideas and reducing the strain on hands. Significantly, Tambourine outperforms standard dictation tools by utilizing advanced AI techniques to cleanly format speech, removing filler words, adding punctuation, and customizing terms through a personal dictionary. The platform's flexibility is amplified by the ability to select various speech-to-text (STT) and language model (LLM) providers, configure prompts, and adjust formatting rules to suit individual needs. Additionally, its open-source nature not only promotes transparency but also enables endless customization, fostering a community-driven approach to voice dictation technology. With features such as context-aware formatting and voice-driven text modification, Tambourine positions itself as a powerful tool for anyone looking to streamline their digital communication.