Web App to extract and browse through tabular data using Flask. (c) Alexander Hirner 2016, no redistribution without permission.

Municipalities suffer from ill financed public infrastructure (e.g. Flint water scandal). A large part of this problem is caused by lack of transparency in the municipal bond market. To solve that, we extract financial data buried in pdf documents, aggregate them in comparable and tabular form, and thus enable officials to bargain better terms. This type of analysis has merit in many other verticals too. The data (>1000 docs) and NLP talent is there, now we need UI/UX and Open Data enthusiasts!

We use cortical.io / retina SDK API to generate embedding. We use IBM BlueMix to perform PDF to text conversion.

npm install -g bower
pip install -r requirements.txt
bower install
python server.py

Navigate to http://localhost:8000 and upload an example document (see below).

• ./templates ... Jinja2 template
• ,/static ... all stylesheets and media goes there
• ./static/ug/<project_name> ... user uploaded data and analysis files (graphs, json)

see EXAMPLES.md