Executive Summary

Social media has become an important part of the “Marketplace of Ideas”. There is a clear difference between an expert and a troll, a journalist and a bot, between ‘Real news’ and ‘Fake News’. And with half a billion strangers online, it’s not easy to tell who is who.

But we have the solution.

Clairety is a Rating Agency for social media content.Our Artificial Intelligence system tracks the credibility of all half a billion social media contributors, and reports it to you with a browser module that embeds our ratings directly in your social media content. Our fully automated system is completely unbiased and uses no human interaction. We identify bots, spam, harassment, extreme speech and a host of other ‘low value’ content. We can separate facts from opinions, expertise from marketing, and the ‘real story’ from conspiracy theory. We let you know who is risky to believe, and who isn’t,providing real economic value to Social Media consumers, publishers, advertisers, and those hoping to better understand social media and the marketplace of ideas.

Why Credibility?

Credibility is an intuitive principle that all of us use in our day to day lives. By bringing this face to face concept online, we make the social media experience better resemble in person exchanges.

Real Economic Value

Bots, automated accounts, trolls, and harassers are all clearly identified in our system, providing real economic value to both Social media users and advertisers.

Truly Unbiased

Our AI has no views of biases of its own, and involves no human interaction. It creates a purely relative ranking where every contributor is measured only in relation to all the others.

Commercially Proven

The core logic of has been proven successful in the financial markets, reading newswires for program trading systems.

Totally Independent

Social Media platforms need to treat all users fairly. As a totally independent company our only incentive is to measure the differences between them fairly. We will never make content or sell content. Our only goal is to tell you what is safe and reliable.

Supporting Crypto Traders

Our Crypto trading dataset will tell you who to trust in crypto Social Media and who not to. Letting traders see the difference between real information and spam, and increasing the ‘credibility’ of the entire crypto ecosystem.

Pangaea Screenshots
Divider Background

The Crypto-ecosystem’s dependence on social media is absolute. Our crypto-specific dataset will analyze lower volume specialty social media platforms like Cointalk which are essential to understanding the dialog in the crypto space. Specifically designed for quantitative trading with one of our partners who specializes in Crypto, this dataset can tell you who to believe, and who not to, and we’re confident it will contribute meaningful alpha to your trading process.

The crypto space needs a credibility boost. We’re here to help provide it.

Meet The Team

Our team has nearly 50 years of combined experience in executive management and data science, and have been working with NLP and large datasets for over a decade.

Tom Costello - CEO/CTO
Tom Costello


Tom Costello started his career at the New York city headquarters of JPMorgan where he worked as a quantitative analyst and developer on the trading desk designing derivatives and pricing systems for institutional clients. He moved to the Hedge Fund side taking a position in Research at Moore Capital


Randall Brown - CFO
Randall Brown


As Head of Rates, Sales and Trading for Mitsubishi UFJ Securities Mr. Brown was responsible for all aspects of funding, risk management, and oversight of the bank’s US fixed income trading operations. He was a member of the Management Committee, Asset/Liability Committee, Risk Committee


We're Hiring - view our list of available positions

click here to view our list of available positions

FAQ Background


Here are some of our most frequently asked questions. To see the full and extensive list, please click here.

1. How does it ‘actually’ work?

It’s no joke. We have an artificial intelligence system that reads public social media comments and ranks the author based on ‘credibility’. We already have a working prototype connected to Social Media producing stunning results. And a system with the same basic logic has been profitably running for years in the financial markets.

2. Who else is doing something like this?

Rating agencies like ours operate in a number of industries. One of the longest running, and in our opinion the most successful example is Underwriter’s Labs, which has been rating the safety and reliability of electrical equipment since 1894, and now operates in 106 countries and has thousands of employees.
Thomas Edison was running GE in 1894 and saw UL’s independent rating as a means of preserving the incentives necessary for gaining consumer confidence, while reducing the threat of government regulation. So rather than crushing them as he easily could have, instead he decided to partner with them and improve his industry overall.
We’re the first to attempt this in the social media space, but it’s a well established economic niche. And with a decade of experience with the technology in the financial markets, we believe it’s up to the challenge.

3. What if I want to know if a specific news story is ‘Fake’?

Fake according to who? Very little media content is totally fictional.
Most ‘fake news’ is actually describing the gap between the bias of the publisher and the bias of the reader. Far left people call right leaning news ‘fake’ and far right people call left leaning news ‘fake’. What we do is map the degree of ‘bias’ in either direction, so we give you an idea of how many people think a news story is fake overall.
No one in media writes a story popular with Rachel Maddow fans on Monday and another that’s popular with Alex Jones fans on Tuesday. Because it’s closely tied with their psychology and the values they hold most dear, people generally lean one way or another. What we give you is a statistic which tells you how hard they’re leaning right now, not which direction they’re leaning.

4. How do you know something is a Bot?

For all their attention in the social media space, most bots aren’t very sophisticated. Most of them are fairly simple programs which repeat stories in mutually exclusive media bubbles in order to get more clicks. Some don’t even go that far and only exist to boost the ‘follower’ or friend counts of a Social Media user.
Since our system has no biases and is reading everything, we aren’t limited by a bubble. So it’s fairly easy for us to tell who is producing original content and who isn’t.

5. What if I don’t want to be scored by you?

We’re only interested in the quality of the public ‘marketplace of ideas’. If you don’t want to be rated by us, then don’t make your comments publicly. Everyone is entitled to their opinion and we respect that absolutely, whatever the opinion is. But if you make that comment publicly, then the public is entitled to review it and draw their own conclusions about you based on what you say.

6. What if a person’s social media is ‘private’?

We don’t read ‘private’ social media comments and never will. We firmly believe that what you say in private is no one’s business but your own.

7. Google Facebook and Twitter are multi-billion dollar Corporations with advanced expertise in NLP. What makes you think they won’t do something like this and cut you out?

Well our US patent gives us some protection, but our real advantage is our independence. All of the major social media platforms have been coping with very credible accusations of bias in their judgement of ‘fake news’, and this has done major damage to their brands.
We believe they’re caught between a rock and a hard place. Torn between their business model which demands that they treat everyone equally, and being the speech police which suits them less well.
It makes more sense for them to partner with us than to do it themselves, just like Thomas Edison did with Underwriter’s Labs in 1894. It’s better for them, and it’s better for us.

8. Don’t you think the social media companies will see you as moving into their territory?

We certainly hope not. That’s not how we see ourselves. We want to make things better for them not worse. We can reduce the potential threat of litigation, let them focus on their true business model by treating people equally, provide incentives that will improve the quality of news media, reduce cultural polarization, and we can do it all without it affecting their bottom line.

If you see them, by all means tell them to give us a call.

9. What makes you think you’ll be seen as an objective source?

The basic logic of this system was developed and proven effective in the financial markets; an environment that instantly and ruthlessly punishes anything except absolute and total objectivity.
We don’t take a side on any subject. And we believe our results will make our objectivity very clear. Our testing data from our working prototype was completely persuasive and matched human intuition for credibility very effectively.
We also found that we have people who score both high and low on the political left and the political right, so it clearly isn’t about politics for us. Our AI doesn’t have a view. We only want to give our users enough control over their social media to avoid wasted time and aggravation from bots, spam, harassment and other low value information. How low? That’s up to the individual users.

10. Why does [insert person here] have such a [Low/High] score?

All we look at is what people say online. If someone has given an interview on TV which caused you to have a particular feeling about their credibility, we won’t see that and it won’t affect their score.
But what you say is a component of who you are. So especially when it comes to public figures, we believe that over time, a person’s credibility can be accurately assessed by the totality of the things they say online. If they really have extreme ideas then eventually they’ll say something about it which our AI identifies as extreme.

11. I think [insert person here] is extremely credible, why do they have such a [Low/High] score.

In the broadest terms, we tend to think the people we agree with are more credible than those we disagree with. But this is a reflection of our own individual biases.
Our system has no view of its own so it doesn’t do that. It forms its ‘viewpoint’ from the totality of the statements of others that it reads online.
You can think of our statistic of a reflection of that broadest view. You may find the person credible or not, but there are a whole lot of people who might not share your viewpoint who feel very differently about it. We treat all those opinions as valid, whatever they are.

12. I can draw my own conclusions about ‘Fake News’ what do I need you for?

Your conclusions on what news is ‘fake’ and what is not is a reflection your specific viewpoint, which is exactly as it should be. You’re perfectly entitled to your view, whatever that view is. But there are a lot of people on social media and they don’t all see things the same way. Maybe there is ‘real’ information in the news that you think is ‘fake’, and it would benefit you to read it. Or maybe all you want to know is ‘how fake’.
If you’re interested in how people who don’t share your viewpoint see things, ours will be better metric of ‘fake’ vs ‘real’ than your individual view will. And who knows… maybe you’re interested in how credible you seem to others? We have plans for products delivered to HR managers, to advertisers and a whole host of people who interact with social media professionally.
Besides, the basic product is free so you might as well take a look.

13. What if your rating is just ‘wrong’?

Our ‘credibility ranking’ is a relative measure, so it isn’t really possible for it to be completely wrong. What we’re doing is akin to measuring people with a single ruler and reporting their relative height differences. If we’re wrong at all, we’re wrong in the same amount and to the same degree for everyone. And since we only describe the differences, THAT will be right.

We already know there is some portion of information that we miss and a correction for that is built into the system. We’re never going to be 100% right on 100% of everything. But reducing that error to its mathematical minimum is our biggest technical priority, and to the degree we succeed, we’ll make our product better.

14. How do I use your crypto data to improve my trading returns?

Our CSMF dataset is a tool for professional traders who use data and models to make assessments of the markets. Even on our very small team we have over 50 years of experience in building and developing profitable trading models at some of the premier financial institutions in the world, so we know what goes into the best of them. We know how to obtain an ‘information advantage’, and we believe we can help provide one.
To professional traders this data will be fairly intuitive especially if they’ve worked in the fiat trading world. For them, a detailed description of the data and its formation process will be released along with the dataset, and future changes to our process will be tracked over time for historical comparisons and backtesting.
With that said, based on our experience in the space we are completely confident that this data can be used to obtain ‘alpha’ in the blockchain trading world, and are devoted to preserving it as a value added dataset.
This article will explain to you what a small part of the problem is - buzzfeed.com/ryanmac

15. If this data is so good why don’t you just trade Blockchain yourself?

Trading profitability is limited by liquidity. We’ve seen evidence here in New York that in 2018, a great many more fiat financial institutions will be entering the crypto trading space. We believe that a product like ours in that space adds a great deal to the total ecosystem by demonstrating that the market is maturing, by fighting fraud, abuse, and improving the ecosystem’s credibility overall.
But Clairety isn’t just about the blockchain world. The projected revenue from our dollar denominated social media products has the potential to be in the hundreds of millions of dollars a year, maybe even more. It also holds the potential to improve the online world in many ways that can’t be quantified. We like the idea of improving the world at least as much as we like turning a profit. Doing both has an irresistible appeal.

16. Most blockchain traders already have their own credible ‘sources’, so why do they need your data?

As we said, our crypto trading dataset is a tool for professional traders using models to trade. Whatever you think the blockchain trading world is today, in a year we think it’s going to look very different.
The traditional finance world has finally discovered blockchain in a big way and this is a tool much like the ones that they already use in other markets.
Automation and ‘data driven trading’ have been on a long march through the institutions of the trading world for decades. Given the extreme technical savvy of the members of the blockchain ecosystem, we believe this market is going to leapfrog most of that effort. ‘Quant’ trading is coming to crypto. And when it does, our data is exactly what they’ll be hoping to find.
We’ve been buying data like this to support our own trading models in the fiat trading world for years. One of our closest advisors has been selling data and systems in the financial space both to us and to other like us, for decades. We all see this rapidly evolving market the same way.

17. Can it be hacked?

The short answer is that we don’t think so no. We confess that our data may be a potential target, but that’s the whole point of including a blockchain specific access methodology. If you trust the security of ETH, then you should be able to trust our data integrity.

18. Will other people know I’m using it?

Nope. And that goes for all our products. Only you and we will ever know how you use our system, or if you’re using it at all. We won’t even tell the social media providers.

19. How credible is “Donald Trump’s Twitter”?

Everyone asks us this eventually. Whether you love him or hate him, more than anyone else in public office, Donald Trump has “a unique way with words”. When compared to other more traditional politicians he scores in the lower end of the spectrum, but he’s not the lowest by any means. And there are millions and millions of people who score lower than him from outside politics.

All told I’d say he’s about where people would expect him to be. But his score changes every time he sends out a tweet, and he sends out a lot of them. To get the whole picture you’ll have to sign up for the free product when it’s ready, and see for yourself.

20. What about the score for my [least/most] favorite media source?

What really struck us was how closely the media producers all scored to each other. They’re all bundled up in the same tight little group CNN, right beside FoxNews, beside the WSJ, beside the NYTimes. They change order fairly frequently, but they’re all about the same as each other most of the time.
What that tells me is that even though they all have different views, they are all responding to the same economic incentives in terms of how much they editorialize their content.
The fact of the matter is that when incentives are considered, the people who have the most to gain from the success of our product will probably be the established news media. Right now, with all information being treated as exactly equal in value, a click for a bot is worth just as much for a click on the WSJ.
But when there is a number like ours out there describing the quality of content, those people who can produce higher quality content will have an incentive to do so. The mainstream media has huge infrastructure and resources. Once they have an incentive to produce higher quality news, I think they’ll be happy (maybe even relieved) to.

21. Do you think social media is going to be regulated?

We hope not. We prefer free and open, especially when it comes to online speech. But there is reason to suspect it might be.
In June 2017, The US Supreme Court ruled on a case called Packingham v. North Carolina where Justice Kennedy described social media as “a part of the public square”. That had no specific force of law, but it immediately inspired several lawsuits trying to restrict the government and social media companies from ‘banning’ individuals. That’s a long way from banning speech, but it’s along that road. And we can see how it might get there from here.
We think it’s a bad idea to let the government get further involved in the restriction of speech, and we think we have a way for people to get the protection they feel they need from online harassment and other nuisances, without resorting to more regulation.
To us, more speech is the solution to bad speech, but only if people can tell the difference between good and bad speech.
Not all ideas are equal. And we aim to give people the information they need to be able to tell the difference between high quality and low.
This article will help to explain our thoughts on this - cnbc.com

Contact Us

If you'd like to get in touch with us for any reason what so ever, you can do so by either calling or emailing us using the details below. Alternatively, use the contact form and we'll be sure to get back to you as soon as we possibly can.

We're located in New York City