Web interface user wants to:

• browse available dataset on a web site
• select some time series to compare
• download selected data locally

This user is likely to use data browser at http://macrodash.herokuapp.com/.

R or pandas users with experience in FRED or quandl would want:

• a clean dataset with latest data from different sources
• download this data on a local machine (in pandas or R)
• quickly draw some charts (like one below)
• develop explanatory/forecasting models and share them as IPython notebooks.

Example: read official daily ruble/usd exchange rate from start of 2017

import pandas as pd

def read_ts(source_url):
    """Read pandas time series from *source_url*."""
    return pd.read_csv(source_url, 
                      converters={0: pd.to_datetime}, 
                      index_col=0,
                      squeeze=True)

er = read_ts('http://minikep-db.herokuapp.com/ru/series/USDRUR_CB/d/2017')
assert er['2017-09-28'] == 58.01022

Click here to see the same data in browser.