This is a pet project I've been hacking away at for a bit. I don't expect the audience to be huge, but I'm just wondering if anyone would ever use such a service, or even pay a small amount of money for it. I don't think it would be worth making a company just for this one service; I'd likely run it as a one-man side operation, unless it became popular quickly. My description might sound a bit vague, so please ask me to clarify any points if necessary.
Full description here, as it couldn't all fit into a HN post: http://textuploader.com/d3fk
Technical
A nice, simple API and web application for webmasters to view, search, and manipulate visitor device & network fingerprints and signatures. Various techniques are used to facilitate enforcement of bans, and detection of suspicious activity, registrations, and submissions.
For end users
A security service targeted towards web admins of every kind, specifically to combat bots, spammers, fraudsters, and anyone with malevolent intent. Site owners would register their domain on the service's website, and be given a snippet of Javascript to place in the head of each page they serve.
Customers would log into our site to see recent visits, anomalies, and other passive information. They would have a robust querying interface, and would also be able to tie device fingerprints to user session variables, like username. They will also have the ability to make complex rules and bans.
General
Customers can search by IP address, CIDR mask, location, date/time, or by the fingerprint itself (either a composite fingerprint or one of many more specific ones, to vary the false positive and false negative rate); a more basic interface will be available for those who have no need for fine grained searching.
I list some concerns and similarities to other services in my full description.
I'll gladly take any suggestions or criticism. Thanks.
My second thought was about privacy and that the fact that you might be selling aggregate information about visitors of your clients' sites. I am glad that you addressed that as well.
I am concerned about your attempts to distinguish yourself, especially by your suggestion that it will be "more hands-on, less machine-driven". Since I would want to stop fraudulent/malicious behavior while/before it happens, and since I can review user actions after the fact by myself, I am not sure whether your service would be the best fit for someone who might pay Sift or a similar company.