You are an expert in the New York transportation system and are a senior developer, highly experienced in HTML, CSS, JS, etc.
Here is some context: NYC Open Data is the city's main public data housing system, built on the Socrata platform. It is open to the public and uses a REST API (SODA) that you can query and filter with SoQL. Open data has stayed virtually the same since it was opened, and could use some upgrades.
1) First Issue: Open Data lacks a quick data summary tool. Right now if I open a dataset on Open Data, I get a raw table of rows and a column list, with no real sense of what's actually in it. I want a quick scan / dataset summary feature and something that lands on a dataset and immediately gives you a snapshot of it. Ideally it could also let you jump to a few random points in the data so you can actually peek at real examples instead of just reading column headers and guessing.
2) Second Issue: The built in filters are quite slow and dated: The current filtering system in Open Data allows for filtering columns one field at a time. I want to have better more flexible filters, like for example letting users draw custom area on a map for datasets with geospatial information. I also want to be able to layer datasets and filter from the combination.
3) Third Issue: There are no good ways to browse datasets by location: Open Data's databases are organized by category, or can be found through keywords. The keyword search can be useful but is overall not a perfect experience. I want a location-based browser, so that I can pick a neighborhood or borough and see which datasets actually have info covering that area.
Help me design and code (HTML, CSS, JS,) these features into the Open Data platform. Walk me through your approach, then lets build it out.