- intelligent filtering and credibility from TrueKred


When people post content today, there are few automated filters for profanity, irrelevance and hate and even decent content has no uniform, meaningful way to gauge it's credibility (unless the reader knows the author well). When users leave feedback on ecommerce sites or reviews on local sites, there is no way to understand just how well they know the people, places or subject matter they post about. Truekred aims to solve this problem by doing 2 things:

  1. Filtering content for profanity, trolling and irrelevancy and

  2. By providing a scoring system for users, whether they post as themselves or choose to be anonymous, by analyzing the content they post and the responses they get from the community, across their online or mobile footprints. Using advanced Social Graph analysis, Machine Learning, Natural Language Processing and Sentiment analysis, a picture emerges of things people write about, how well they are regarded for it and how well they know both, the entities, as well as the subject matter they write about.

    As users assimilate Truekred scores (these scores are real-time and will constantly update) and apps and sites start using these scores, a standard mechanism will evolve for the community to value and consume content. Whether a user is leaving feedback on a seller on ebay or a review for a product on Amazon or a store on Yelp or an anonymous post about some breaking news on an app like Secret or Rooms, consumers should be able to look at a score attached to that content and then judge how they want to treat it.

    Getting to this point is possible because the Truekred system has the ability to understand the authors social graph where possible, identify and understand what is being written, how it's being received and even to some extent, deciding if some of the content is a violation of policy, someone's rights or even portends to abuse.


Signing up is easy - create a TrueKred account online and connect your social graph. The system will start analyzing your network, the content you have posted and the response you have received for it. Quickly, TrueKred will start generating topics of knowledge, scores for each of them and a graph of your network.

Apps and sites that support TrueKred that you authorize will start displaying your real-time credibility scores as you use them.



Quantity isn't everything. Infact it is often very deceiving. If you value a user as a subject matter expert because she has a few thousand comments on a post about Adventure Travel to Asia, you could be missing the real gist of those comments - perhaps a withiring critique of the experience and the suggestions made by the author. With TrueKred, it's all about the quality of the content and how well it is received by the community rather than just how many views it has.

You want to engage with quality audiences, not unhappy readers or often uninterested trolls.

You can subscribe to a corporate TrueKred account and identify high value users by topics.


Add TrueKred to your websites, apps or intranet via our easy to use API. You can use our filtering API to catch undesirable content and you can also use our global TrueKred scores for your users or use the scoring engine just within your domain.

The TrueKred API is releasing soon and we are adding partners gradually. If you would like to see how TrueKred can bring credibility into your service, talk to us and we'll show you how easy it is.

About Us

We are active social media users, who absolutely hate the fact that there is no good way to decipher the good from the trash, the real from the baloney, the authentic from the drivel anonymous trolls dish out on the web - in posts, in comments, in discussions, in reviews, in feedback, on ecommerce sites. You get the picture. We chanelled this pain and used our backgrounds in machine- learning, natural language processing and sentiment analysis to come up with a decent way to build and find credibility online - whether anonymous or as yourself.

We are San Francisco based and sincerely hope to make this place a more pleasant web.