Orato
Practice speaking with AI.
159 followers
Practice speaking with AI.
159 followers
Pick a drill, speak for 30 to 90 seconds, and get scored on pacing, fluency, vocabulary and coherence. Every filler and long pause lands on a transcript you can read back. On an iPhone with Apple Intelligence it runs on the device, with no account.
Free Options
Launch Team
Render Sandboxes: a computer for every agentIsolated compute, persistent state and a real filesystem
Promoted
Congrats on the launch :)))
On device with no account is the reason this can work. Nobody practises sounding bad in front of a server they don't trust, and the whole thing depends on people being willing to do a terrible first take. The scoring is where I'd be careful. Pacing and fluency are measurable, coherence is a judgement call, and if that one is wrong even once people quietly stop believing the other three.
Thanks for this, Asad. Your point about one bad coherence judgment undermining the other feedback makes sense. A score on its own leaves too much unexplained. I'd want the feedback to point to the part of the answer it is questioning, so the speaker can decide whether the criticism is fair. What would you need to see alongside that score to find it useful?