Merge pull request #87 from scale-vector/intro

first draft of intro

docs/website/docs/intro.md

@@ -4,10 +4,27 @@ sidebar_position: 1
 
 # Introduction
 
+![dlt pacman](/img/dlt-pacman.gif)
+
 ## Summary
 
+`dlt` is an open source library that you include in a Python script to create a highly 
+scalable, easy to maintain, straightforward to deploy data [pipeline](./glossary.md#pipeline).
+Once you set it up, it will then automatically turn JSON returned by any 
+[source](./glossary.md#source) (e.g. an API) into a live dataset stored in the 
+[destination](./glossary.md#destination) of your choice (e.g. Google BigQuery).
+
 ## Who should use `dlt`?
 
+Anyone who solves problems in their job using Python should use `dlt` (e.g. data scientists, 
+data analysts, etc).
+
 ## What does `dlt` do?
 
-## How does `dlt` work?
+`dlt` is used to automate fetching data for others, copying production data to somewhere else, putting data from an API into a database, getting data for dashboards that automatically refresh with new data, and more.
+
+## How does `dlt` work?
+
+`dlt` extracts data from a [source](./glossary.md#source), inspects its structure to create a [schema](./glossary.md#schema), normalizes and verifies the data,
+and then loads the data into a [destination](./glossary.md#destination). 
+You can read more about how it works [here](./architecture.md).