Automated Transcription Pipeline

Over 72 hours saved | Bulk speech-to-text file and metadata record creation

Project Specs

Created low-code, open-source automated transcription pipeline to generate speech-to-text files and metadata indexing

Demo

Video short includes a demo of the pipeline. For more info, including repository and download instructions, see https://github.com/catatwork217

Speech-to-text automated transcription pipeline and end-to-end data management

  • OpenAI whisper CLI

    - Early iterations integrated Shell scripts and OpenAI Whisper CLI in Bash terminal to enable logging, speech-to-text transcription on-demand, resulting in a text, caption-ready and JSON file.

  • Pipeline enhancement in python

    - Enhanced pipeline to manage end-to-end transcription and record creation in Python, generating transcription record in SQLite with word count, keyword metadata, and logging.

  • Data management in the pipeline

    - Added data management capabilities like duplicate validation, exception tracking and record adjustment.

  • Advanced data management in postgreSQL & pgadmin

    - Integrated PostgreSQL dB end-to-end transcription data management into Python script, enabling server config, table records word search and analysis.

Back to AI & Automation Projects
Next

Process Automation - Badge Format Analysis & Creation