Interactive Lab

Hands-on exercises to build and test a voice agent. Follow these in order for the best experience.

Exercise 1: Local STT Pipeline

Set up whisper.cpp and transcribe audio in real-time from your microphone.

bash

# Build whisper.cpp with streaming support
git clone https://github.com/ggerganov/whisper.cpp
cd whisper.cpp && make
bash models/download-ggml-model.sh base.en

# Test with a file first
./main -m models/ggml-base.en.bin -f test_audio.wav

# Stream from microphone
./stream -m models/ggml-base.en.bin \
  --step 500 --length 5000 -t 4

# Experiment: try different models
# tiny.en (fastest), base.en (balanced), small.en (most accurate)

Satisfied with your knowledge? Continue to the next module.

Continue to Social Engineering

Loading…