Develop a prototype crawler + query builder + interface + file host + download

The idea behind the service is to capture with a very limited delay, whenever calls for bidding are done in client’s websites. Since there is no ‘structured service’ or ‘single’ marketplace for that, the crawler will have to navigate by itself to find these results. Tender’s Sniffer is a data crawler designed to locate specific results with parameters and operators. The idea behind it is to establish sources of websites and perform searches for specific files within these ‘lists for websites’. The searches should be ‘limited’ to prevent flooding, but continuous with a specific timespan between repeated searches. Whenever there’s a match, this match will go for evaluation of the user, to judge it as ‘good/not good’ and to download or not. This is only a prototype to validate some assumptions. The interface, infra-structure, speed and security are not the main issues now, but the core of the system is. Please read the attached file to understand what I am exactly looking for.