This workshop introduces a Google Colab toolkit for collecting, transcribing, and combining speech and live chat data from recorded YouTube and Twitch video streams. Participants will learn how to retrieve video metadata, download live chat logs, generate automatic speech recognition (ASR) transcripts with faster-whisper, and integrate multimodal interaction data into unified corpora for analysis. The workflow also demonstrates data structuring, metadata handling, and export formats for downstream computational and corpus-linguistic research. The notebook provides a practical foundation for studying livestream interaction, digital discourse, and multimodal online communication at scale. Workshop participants will utilize Google's Colab service to run Python code on a Jupyter notebook. To use Colab, a Google account is necessary (accounts.google.com).