Improve usefulness of Vectara Answer

[cjcenizal]

CC @ofermend

Problem

1. Unclear message in "About". There's a lot of good info in this section but it's hard to find the part that actually describes what Vectara Answer can be used for.
2. Quick Start has too much friction. It requires Docker, python3, pip, pyyaml, npm, and node. It requires ingesting data into Vectara, and implies that you have to do that with Vectara Ingest.
3. Tight coupling with Vectara Ingest. The Quick Start expects the user to do the Vectara Ingest Quick Start first.
4. README contains too much information. It's hard to find what you're looking for on such a long page.

Proposed solutions

Enable configs to work out-of-the-box

We can add the correct customer IDs, corpus IDs, and API keys to each config.yaml file so that they work out of the box. They keys are query-only and the data isn't sensitive, so this won't be a security concern.

Rewrite "About" to focus on the purpose of the repo

Vectara Answer is an open source React project that enables you to quickly configure GenAI user interfaces, powered by the Vectara Platform's semantic search and summarization APIs.

Simplify Quick Start

Let's simplify Quick Start to getting a config running locally, using just node and NPM. This will remove the dependencies on Docker and Python. This will require some engineering to make our local dev process compatible with the config system.

Extract configuration and deployment content

Extract this content into CONFIGURATION.md and DEPLOYMENT.md files. Link to these from the README.

Present Vectara Ingest as optional

When the topic of ingestion comes up, mention that if folks need help getting data into Vectara they can check out our ingestion docs or use Vectara Ingest.

Add screenshots to README

We should add screenshots of the resulting UI to the top of the README so folks understand the what this project will produce for them.

Add visual diagram of UI configuration

This will help folks understand the visual impact of making configuration changes.