Video-based web application

I have a partially developed, golf-related web application that is written the R programming language and Shiny application (partly because they were supposed to be quicker and/or easier for a first version, though this has turned out to be anything but). The finished application certainly does not have to be in R. In fact, the partially finished application runs extremely slowly in that programming and might ultimately have to be rewritten anyhow (although there are some nice visuals possible within the application and I do not know if they can be comparatively duplicated in other languages). But someone that can at least skillfully decipher and transpose that language efficiently would seemingly be helpful to finally get this application up and running as soon as possible. The previous guy said he could build the app from scratch in about five days and has fallen far short of any success after almost a year now. I cannot tell if the programming language(s) have gotten in the way of success at all, if the guy was just not skilled enough, or if perhaps it has been some of both. Anyhow, users start by recording their own videos of their golf pre-swing motions (which could last upward of 15 seconds in length or so for some golfers). A selected video is uploaded to the application. The original video format (which could vary) is reformatted to MP4 (I think) by the application, and frame numbers are added and displayed within the reformatted video (time code could be used instead if more efficient). The reformatted video is then visually analyzed essentially frame by frame by the user. (This can be done efficiently in some browsers and not others, so an application option exists for downloading the reformatted video for analysis in a discrete local video player). One or more shorter segments of the video are selected by the user to be further analyzed by the application. The determined segment(s) is then input into an analysis section of the application and certain visual and specified mathematical parameters are output. What needs to be specifically and accurately tracked and analyzed is the golf club movement. This has been done thus far in the application by a color-tracking computer vision program, with the program marking and tracking two colored markers originally placed at known positions on a club before recording a video. An auto-placement feature of tracking markers by the computer vision program is great and needed, and yet it has never yet been 100% correct in placing tracking markers when reviewing its accuracy, even for video segments comprising as few as ten frames (a reviewing feature that is included and needed in the application). But a strictly manual marking procedure is problematical also, as even very short (time-wise) video segments could have high frame rates and be much too cumbersome to mark manually. So some type of "hybrid" tracking system would seem to be in order at this time, consisting of an auto-tracking system but with a supplemental feature option to manually alter markers upon reviewing their auto-tracking placements as subsequently needed. The past contractor has been unable to accomplish this club tracking efficiently and successfully as the first fundamental milestone, so we have been largely unable to even get to the analysis and output parts yet. He has since stated that he would try to work on a faster and perhaps more accurate tracking method more along the lines of Artificial Intelligence, and I am open to that, but I have just found out that he has really not been working on it at all. This needs to be finished and up and running as soon as possible. I need help with finishing the application, and when successful, deploying it online. Hourly or project quotes can be considered. Thanks.