Skip to content
All projects

AI · Real-Time· 2026 · Creator

Echo — Real-Time AI Speaking Coach

Real-time AI speaking coach with live transcription, filler-word detection, eye-contact tracking, mock interviews, and Gemini-powered feedback reports.

  • Live transcription with filler-word detection
  • Eye-contact tracking during practice sessions
  • Mock interviews with Gemini-powered feedback reports
TypeScriptGemini APIBrowser Media APIsReal-Time

Context & Goals

Echo turns the browser into a speaking coach. You practice — a talk, an answer, a mock interview — and it listens and watches in real time: transcribing live, flagging filler words, and tracking eye contact. Afterward, Gemini generates a structured feedback report.

What it does

  • Live transcription while you speak, with filler words ("um", "like") surfaced as they happen.
  • Eye-contact tracking so you know when you're drifting off-camera.
  • Mock interview mode for structured practice.
  • Gemini-powered reports summarizing strengths and what to fix next.

Up next

PulseMQ — Distributed Message Queue in Go

A distributed message queue built from scratch in Go — an exercise in queueing semantics, concurrency, and distributed-systems fundamentals.

View project