feat: add starter kit for Elasticsearch using PDF and Office Documents

[sne3091]

This PR adds a new conversational search guide using Elasticsearch for PDF and Office Documents. This is Part 1, which is a low code approach. I'm also working on adding some content for a high code approach using Python libraries

Signed-off by: sneha.srinivasan1130@ibm.com

[sne3091]

@jwm4 Still working on referencing out to a separate pre-requisites doc for docker , other comments that we dicussed have been addressed

Also, the actions.json file in this PR is now the same as what @zach-shu has in his PR with the minor modification he made to my previous proposed change on the same file( I tested it on WA with this guide as well again ). I kept here just because we can merge whichever is ready first and either one of us can remove it from our PRs whenever that is done to avoid duplication

[sne3091]

I also have a general question about docker vs docker-compose. Would using docker simplify the setup? If it is just one command to start the docker container, I would prefer to use docker.

@zach-shu I kept compose from the point of view of organization in the long run if we were to ask a user to startup multiple containers like Kibana, Enterprise Search etc. , then we could find a better way to set things up such that they would only have to run 1 command using a single file. Its not included in this guide yet however if a user wanted to visually inspect , then the kibana config can also be added into the compose YAML to stand up both containers together

I would still prefer to use compose nevertheless , since the existing guide accounts for that in setup and also boxes the whole fscrawler setup into a single directory of the configs/logs etc as opposed to using docker where its a little bit more disorganized

[sne3091]

Can anyone with write access please merge? Thank you!