Using torrent consists of multiple different parts. Some of which depends on your actual implementation. Parts you would need to have a look at include:
- Torrent files
- Magnet URIs
- Clients
- Trackers
- Peer exchange technologies like DHT
Torrent Files
A torrent file contains all the important meta information about the torrent. If you look at the Wikipedia article you can see a decoded example of a torrent file. Note that you have a list of files and the hashes of pieces. If you want to add a file to this structure you will need to modify both files. In this regard the Wikipedia article on the protocol might also be interesting. If you are using a torrent file you can't just add some data to it. If you do change this information you're going to change the hash value of the information section which is used to identify the torrent. As such you will create a new torrent. See also What exactly is the info_Hash in a torrent file.
Torrent Clients
Torrent Clients are part of the user side of using torrents. There are various clients available and some of them offer to automatically pickup torrent files from directories. An example here could be that you have a single file per backup that the other clients pickup. As torrent files are quite small it might be possible to have a script, a common share where they are stored or you might email them later on.
Magnet URIs
Again the Wikipedia article has quite a bit of information but is rather technical. Especially since it's not specific to BitTorrent (Magnet can be used for more). A good explanation is offered in How do BitTorrent magnet links work?. It's essentially the same as having a torrent file, it's just that you start with the info hash and need to find a peer who can hand you the actual torrent file.
Trackers
Trackers are rather dumb pieces of software. They only know which peers are interested in a particular info hash and share that information with other peers that are interested in that info hash (that's the basic idea). Piratically there is more to it as often a tracker would also offer an interface and/or might implement strategies to limit who can download. An example for a tracker software would be opentracker.
Peer Exchange Protocols
The technologies (mainly DHT nowadays from what I know) aim to replace trackers. Trackers are a single point of failure after all. So the idea (also describes in the magnet link question) is that you have a swarm that replace the tracker. But essentially it's the same process that is necessary.