Write some Software

Hi, I think I have a project you may find interesting. It combines api web development, speech recognition, and could lead to a position building a call center. I'm looking for a simple website built that allows repeatable tests of any 1 minute (or less) of recorded audio from telephone input, to test the quality of various user-defined solutions, for speech recognition quality. This would include api solutions - I have a list of a dozen speech recognition companies I would want to include. This would be a useful tool for me, and I believe it would also be useful to others. I'm surprised there isn't a site that already does this. I don't want to deal with applicants that haven't done this kind of work before. If you are going to pursue this job, please make your relevant experience clear. Also, I am hoping I can find someone that can easily accomplish far more than this simple introductory job -- someone with experience with freeswitch or equivalent, and that has experience in building call centers. Here's what I'm looking for for this simple introductory job: In the testing website you produce, the user would be able to go to the test site, and choose two things: 1) which of the 12 api speech recognition solutions they want to test, and 2) produce a recording in one of 2 ways: by clicking a link to begin recording of a 1 minute or less audio sample with user microphone, or dial a 1800 number to begin recording a 1 minute sample. The site would access the recording and send it to the api speech recognition of the chosen solution to produce their transcription output, and then would display their 1 minute of transcription output (and any other information they output) on the screen for the user. Overall, I am interested in which technology is closest to reaching the ability to accurately transcribe an ongoing conversation between two people, such as that which would occur in a telephone survey environment. A few thoughts: Also, I would like you to give some advice by thinking innovatively about how to use/access windows speech recognition user training files that may be on the end-users computer -- user calls customer/respondent and provides a link to (audio/video) chat that also gives us permission to access the users windows speech recognition, including the training file. I think a trained windows speech recognition is likely to be better than many of the api solutions. In general, I think I like google, even though it maybe isn't the best, so I would want to make sure to fully explore the capabilities of google, e.g., [login to view URL] The core of this is a test of the speech recognition capabilities of various products. This is for development, so I expect all services will allow use for free to test the products for one minute test calls. After building a way to test each product, I would like you to share your opinion about the quality of each solution. The short term goal is to produce a site that will allow the ongoing testing of the various products, and provide a functionally useful test job for you and I to get acquainted. This can be accomplished, I think, by having you build the testing website described above. This may blossom into ongoing work toward a long term goal -- to put together a full featured call center that also can incorporate social media and transcription, and connect with other programs typically used by a call center, such as survey software (I currently use a great program that runs .net on a windows web server). I run a nonprofit social research company, so there isn't much money, but we have good ideas and a good cause. Thank you for keeping costs as low as possible. Pay is negotiable, although I do not want to pay by the hour (don't know why job is set that way and I can't change it).