LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows
Python ai-agentsdata-engineeringdocument-aigenerative-aiidpjson-extractionllmmcp-serverocrpdf-extractionprompt-engineeringstructured-output
Generating fixtures for tests
Seeding data into the DB can help create fixtures to run tests with a custom set of records that gets reset after every run.
Seeding
In order to seed data into the DB for the first time, we update and make use of seed_data.py to create the necessary records.
Run the following to insert records.
python manage.py shell < utils/seed_data/seed_data.py
Fixture generation
The inserted records/state needs to be dumped to a fixture with the following
python manage.py dumpdata <app_name> --indent 4 > <path_to_filename>.json
Subsequently this can be used to load data (or) reset back to this initial state with the following
python manage.py loaddata <path_to_filename>.json