While this is possible, there are a few things to consider:

1. Evergreen will likely be overkill for your needs, but can be fun to play with.
2. Evergreen runs on Linux, so when it comes to system maintenance (performing upgrades, applying patches, backing up data, etc.), it would help to have some familiarity/comfort with the Linux command line.
3. Evergreen was built by and for professional librarians, who have specialized jargon, workflows, and legacy data formats. For example, metadata for your items would be stored in "bibliographic records" in a format known as MARC. Data in MARC is designated by tags and subfield codes. A title might be signified by tag 245, yet in the library world there are different types of titles, each designated by a different tag or combination of tags and subfields. That said, you could learn just enough to get things to display in your catalog and ignore everything else (though it might horrify a library cataloger), or better, search public servers via Evergreen's "Z39.50" client and download existing bibliographic records and use those in lieu of hand-entering your data.
4. There are single libraries that use Evergreen, but Evergreen was built for handling multiple libraries at once, and you can see that in the prevalence of library selector menus. In general, Evergreen is meant to manage complex libraries, and that calls for some complexity and overhead in turn, things which might not be needed in an application targeting more modest use cases.
5. Items in Evergreen are typically barcoded, though you could just enter their titles if you don't need to track them physically.

Disclaimers aside, probably the first thing you should do is download the staff client and connect to one of the demo servers. There is documentation and tutorials. Play around and see if it might work for you. For running your own server, perhaps the simplest step would be to install Virtualbox and run an Evergreen virtual image with it. Otherwise, check out the download and installation instructions.

Evergreen 2.1 and earlier (circa 2011):

The current OPAC makes heavy use of Javascript (including AJAX) and XML/HTML/CSS. A newer OPAC in development will be moving away from Javascript and toward Perl Template Toolkit for server-side processing. There's also some XSLT being used in places. The current staff client uses the same sort of technology as the current OPAC, but also throws in Mozilla specific technologies such as XUL and XPCOM, and a Javascript library called Dojo. The Evergreen middle layer uses a lot of Perl for the business logic, but also uses C for some services and infrastructure. Certain obscure components revolving around Evergreen may also use Python and Ruby. At the database level we're using PostgreSQL, including stored procedures (in PGSQL, SQL, and Perl) and triggers. Understanding MARC can also be helpful.

See the page for developing code.