Hello! I made this for myself some time ago and I think it is stable enough to share here. This is a super easy way to setup an OpenVPN server with UDP and TCP support along with automatic generation of client .opvn files
We use the app for machine learning purposes. When you get a call that cannot be identified, we ask the user to input the name of the caller.
Additionally, if we identify a call, but the resolution has low confidence, the user will be prompted to vote it up or down. This feeds directly into our identity graph, recalculating in realtime the connections between identities and contact information.
We also use the client app to enhance the existing connections in our identity graph. Each edge in our graph has a confidence level, and by using the data available to the app we adjust and recalculate the confidence levels of our directory in real-time. We never publish data points like photos, notes, etc. from the address book of our users.
- All Contactive API accesses are real time. If we have the data, we will return it right away, if we don't, we will try to find it as soon as we can and return it on the same API request. Full Contact uses the callback pattern, where they notify when they are able or not to resolve the information on a separate call (if they dont have it at the moment of query).
- Contactive not only gives you social network identities, it also gives you aggregated and prioritized data, like the job history of someone, or the list of pictures. This data is returned in the best possible way to be displayed. This saves time from developers because they do not have to make any effort prioritizing each field of the many sources we return.
- Contactive also gives you HTML rendering of a contact, allowing you as a developer to embed it on a iframe. This allows you to integrate it in a web app in a matter of minutes.
The term "big data" stands for the nature of the data Contactive stores, which comes many sources. Most of this data is unstructured and unorganized, and we are able to structure and index it.
Every time we get new data, we have to compare it with the existing set. This is done in realtime..
So, the fact we have 600+ million identities ready to be queried is the result of having a much bigger dataset were all the non-matched-yet data is stored.
We have partnerships with other data companies and we also use the social network APIs.
Everything is stored in a graph and to make the connections between phone, emails and identities we've created algorithms to compare contact information. For this we use signals like name similarity, location, friends in common and many others.
All of this is indexed in realtime and when you do a query to our API you get the most up-to-date data.
The 600+ million records are not just unique records since phone numbers are not necessarily primary keys. One of the challenges of Contactive is about indexing and building relationships between phone numbers, emails and social identities. Every time we collect new data, it is stored into the directory and indexed in realtime.
It is also important to know that we only index public data and also we provide an opt-out option if someone wants to be unlisted.
Enjoy!