Serenade

Serenade Roadmap

Go language support
Ruby language support
Rust language support
Improvements to speech recognition accuracy
Improvements to command understanding accuracy, so commands require less precision
Multi-part add commands, so you can create multiple constructs in a single command
Inferred formatting and symbols when adding code, so you can speak code even faster
New practice mode for learning Serenade commands

Natural English models for dictating prose text
Dictation mode so you don't have to repeat command prefixes
Improved voice activity detection to handle noisy environments
Open-sourced the Serenade Protocol, a way to write your own custom plugins for Serenade
Integrations with the macOS and Windows accessibility APIs, along with a dedicated dictation box in cases where they're unavailable
Expanded the custom command API to include mouse clicks and positioning
First-class terminal support
ML-based command understanding layer