Front-end for time series data

Our company is building algorithms for time-series predictions. We are working towards an MVP and we need a front end. Our backend works in Python and exports relevant results to S3 in Parquet format. We currently have a mock-up front end as the one that can be seen in the screenshot. Now we're looking to build the real thing. You will be given an S3 link with a parquet file containing time series data - you need to have scale in mind, although the file provided will be just a sample file. Done criteria: 1) web service is available at a subdomain on our website. 2) login page with user and password (passwords stored in an AWS Database, as well as file permissions) 3) On successful login: one page with customisable charts, based on the parquet file found in a specific S3 bucket, where the user has access. 4) the charting is easily customisable and user friendly. It really needs to look good! IMPORTANT: 1. Time series are very dense (hourly frequency for +10 years), with +2000 individual time series. 2. The user needs to be able to filter easily on what series will be displayed -- it's not straightforward how to filter on 2000+ options. 3. We need to be able to plot extra information on the graphs, such as confidence intervals and on click text. Currently, we are using Chronograph to chart time series. We need a better solution: faster, more user friendly and easily customisable. We are open to all technologies here, as long as you can convince us it will be suitable for us. Right now we're leaning towards D3/ReactJS. An example of what we want [login to view URL]