Transcript Intelligence Corrections

Current Issues
All files are in the root directory, making it difficult to navigate. You also have incomplete files like 
alert_email_v3.py (which is empty) and duplicate versions like 
alert_email_v2.py, 
alert_email_v3.py, 
email_pipeline.py, and 
email_pipeline1.py. This suggests unfinished refactoring work.
Recommended Directory Structure
transcript-intelligence/
├── src/
│   ├── core/
│   │   ├── insight_extraction.py
│   │   ├── airtable_fetch.py
│   │   └── config.py
│   ├── api/
│   │   └── app.py
│   └── email/
│       ├── alert_email. py
│       └── email_pipeline.py
├── tests/
│   ├── test_airtable_fetch.py
│   ├── test_config.py
│   └── test_web_api. py
├── data/
│   └── insights_final.csv
├── docker/
│   └── Dockerfile
├── requirements.txt
├── README.md
└── .env.example
Key Changes
Start by removing or merging the duplicate email and pipeline files into single, well-maintained versions. Move your CSV data file to a dedicated 
data/ directory and organize your tests in a 
tests/ folder following Python conventions.
Add a README.md to document what your project does and how to use it. Create a 
.env.example file to show required environment variables, and move your Dockerfile to a 
docker/ directory.
For the groups, can we put it somewhere central so that other systems can reuse. It could be on a gist but inside the fvg workspace. Also consider moving the rest of the code there. If you need to be added anywhere, feel free to let me know
Finally, consider adding a 
.github/workflows/ directory for CI/CD if you're not already using it, and optionally use a 
requirements-dev.txt for development-only dependencies.
Review Comment:
GET requests shouldn't have a body, 
so request.get_json()  will return 
None. Change to 
methods=["POST"] instead: 
@app.route("/run", methods=["POST"])
def run_pipeline():
    data = request.get_json()