Develop Mac Application - mini enterprise solution that incorporates acoustic fingerprinting

We're looking to hire a programmer/developer to help our boutique music publishing company create a "mini enterprise solution" of sorts, in order to solve some back catalog management and music search issues. Exposure/experience with any open source “acoustic fingerprinting” or “sonic search” is a plus, as well as having a broad appreciation for music genres, and general music theory knowledge also a plus. Our hope is that a savvy programmer can exploit/tweak on some open source acoustic fingerprinting code, and adapt it to fit our needs, and solve our problem dealing with managing and searching the back catalog of music tracks. As it stands now, our typical workflow for back catalog music search is time intensive and very much like finding a needle in a haystack. It requires multiple humans, to manual search our networked archives and catalogues to find "matching" or "similar" music tracks to the jobs creative brief "reference material.” Essentially, if a client is looking for a music track that sounds like Mumford & Sons, we dig through our back catalog and find tracks that match the vibe and mood of their reference track by ear and by searching metadata. We then present a collection of tracks online via a media hosting cloud based service (similar to Vimeo or the like). The process of tagging files and adding metadata to files in the back catalog is time intensive, we are trying to eliminate this from our workflow. What we'd like our "music search application solution" to do: 1. The application shall allow our music supervisors to "INPUT", via drag and drop, and/or a browse to file location button, the said "reference material" (Usually an mp3 file, or YouTube link [in which case the application would need to convert the YouTube Link on the fly to an mp3]). 2A. The application "listens" or scans the "reference material" with some type of acoustic fingerprinting or other genius wizardry. 2B. If the application can extract tempo and whether the song is in a Major key or Minor key, even better. 3. The application depicts the waveform of the input “reference” file and allows the user to also “trim”, “focus”, or “crop” the file to focus on a specific section of the file. 4. The application has an “accuracy” threshold, a knob or a slider, or input a numerical value 0-100, which allows the user to choose how “tight” or “loose” the search results can be to the reference input. ie. 100% tight would be a perfect match, a duplicate file or a variation of a particular music composition. A 75% threshold would include tracks with same genres, similar tempos, key signatures, and instrumentation extracted from the input acoustic fingerprint. A 25% threshold would include a much looser match, but still be relevant to the input file. All the way down to 0%, which would serve all files in the catalog within the search results. 5. The application then “scans” or cross-references the result from the above functions with our music catalog, or a specified folder directory within the settings menu. ie. If we want to be able to scan the whole catalog, then we point to the parent directory. If we want to scan only record label assets, we need the ability to point the application to only scan that directory. 6. The application then outputs “search results” based on “similar sounding” material to the above input reference file. 7. The application allows the user to play the search results so they can preview them and “cherry pick” the ones they feel are most appropriate for the client pitch. 8. The application then allows the user to drag and drop a file from the search results, and creates a duplicate file upon release of the mouse click. 9. The application can allow the user to save the search results as a playlist for ease of future access to similar searches.