Tint x project and white paper

This is a new project I am working on. One of my first I Have introduced this project to my partners, However I need your expert advice on the project. I appreciate your input, please leave your comment on what you think can be improved, what you think is impossible and what you think is not needed.




    Integrating data analytics and blockchain technology in vetting industrial/investment progress.
    This is a TOKEN + PRODUCT project developed on the blockstack platform aimed at providing better support for token projects and investors and other companies developed on the blockchain technology.

    ICO TOKENS are being created and investor cash are going into businesses that are destined to fail.
    New investors coming into the blockchain technology markets are confused about the technology, where to invest in, how to track company progress pre-and post ICO’s, and where to find potential opportunities, news and extra information.

Currently there is a manual process for solving these issues but not everyone has the time to spend hours researching ICO’s before getting in on it. We take out the time and the work with this project.
We also plan on integrating a tax payment system to assist our clients on taxation on their investments.

First step is to lunch initiate thorough R and D and prototype deployment for this project.
Our product will be deployed within 0-6 months R and D, Prototype Design, basic licensing and product lunch.

Our customers – ICO token issuers, VC’s and Investors, average individuals.
Approximately 30 ICO’s are in pre-opening stages each month. A lot will fail and a few will succeed. We are the filter.

Implementing deep learning and blockchain technology in the evaluation of ICO token issuers. In an ecosystem that has grown in strides within the last 8 years I see an opportunity that no one is paying attention to. Our product will maintain stability and customer confidence in this sector.


  1. 0-6 months R and D, prototype design, basic licensing and product lunch.
    1. 6-12 months re-evaluate prototype with active users re-analyze data and market direction.
  2. 12-24 months lunch of final product.

Devices: Web platform, IOS and Android app.

Hi there!

I don’t really understand how you will evaluate ICOs. You write that you will use deep learning. What should that deep learning model optimize? In other words, how do you mathematically describe a successful ICO?

Also, keep in mind that deep learning models require tonnes of relevant training data. You could just create a huge CSV file describing every single ICO that has ever happened with all the features it had, but that might not be enough. Definitely try though!

How will the users interact with the app? Will the backend run as a dapp so that everyone is 100% sure that your ratings are honest and that no ICO pays you to bump their rating?

Just some food for thought :slight_smile: I wish you good luck with your project, willbe interesting to follow!

Thanks for the comment IVAN, you are practically right. The successful ICO’s will take into consideration the market audience (support from the community) the readiness of direct implementation of that ICO project, similar ICO’s in their space their progress, and the developers weekly activities (Partnerships and integration to other projects),

I am aware that tonnes of data will have to be analyzed and re-analyzed on past and present and new ICO’s introduced in the system, it will also need current information of government/industry regulation and community acceptance (Community Support). Yes the whole project is going to be based on DAPP with a blockchain ledger for storing users information e.t.c.

The idea about a CSV file will be very useful. Transparency is very important, it is very necessary that the accuracy and predictability of this project is among the top 5 of similar platforms or software out there to boost everyone’s confidence in the platform because if an several ICO’s are given a good rating and continuously under-perform’s it is going to show the project in a bad light (This is the most important point to be as accurate as possible).

I sincerely do not know everything that will be implemented right now, I have a lot of Ideas on how to make it better but not too much on if it is feasible to the blockchain technology currently at hand. With you guys as a support community I believe I will not be left in the dark and will ask anytime I get lost lol.

I understand that hard work and dedication plus taking a few things out and tweaking a few others will push the project in the right direction. Not sure how long it will take, maybe a 6 months, 1 year, 3 years or more but I think this project is worth a shot. (If we had this in place the dot com bubble could have been avoided and maybe lost investments could be directed to more promising projects. That is the goal of the project.) Thank you once again.

1 Like

I just created the GitHub repository for this project you can view and contribute here thanks to everyone who shows interest: https://github.com/kncogic/tintx

I tend to look at things more from the business side of things. Here are my thoughts:

  1. Are there really that many ICOs that you need machine learning to help sift through things? Are you solving a problem that is really there?
  2. Is it too early to actually say what is a successful ICO? There are truly only a few coins that are considered successful ICOs right now IMO and they are probably bitcoin based offerings. I think a sense of what is success and what is failure is really going to hamper your ability to collect data.
  3. Is a machine learning even at a state where it can contribute to something with so many unknown-unknowns? ICOs by nature of them being early on in the business’ life, means that there is high amount of uncertainty.
  4. What is your token model? You mentioned that you are going for a token + product. I love ICOs as much as the next guy, but why do you need a token here?

Hopefully these will help you crystallize some thoughts!

1 Like

Thank you Rigario! Your comments are valid and very interesting.

  1. There are about 1100 + ICO’s and tokens active on coin market cap. See: https://coinmarketcap.com/all/views/all/

  2. A little above 300 + more in pre-ICO stages might issue tokens before the end of the year. It is quite early I don’t think too early for this though.

  3. That is the point of machine learning knowing the unknown and iterating continuously, Netflix and a lot of other companies already integrate machine learning and data analysis to pinpoint where they are lacking and how to improve.

  4. Token model is to reduce the actual cost of services correlating user activity to the actual tokens. Full token model is still in planing. Any advice on token modeling? materials, examples? e.t.c.

1 Like

I personally work in a company that has a pretty active machine learning division and the amount of data needed to make a model useful is in the tens to hundreds of thousands. Not to mention all we are trying to do is to understand what a customer is saying a make a prediction on sentiment. In that sense, even if we had 5x the number of ICOs, we may not have enough data to train a model in a useful sense.

Also, when looking at data, its not just the raw number of ICOs that matter. Is what the ICOs have in common. If the ICOs are only going to have a few factors in common, the model produced is not going to be very useful. A good example here is credit score. It takes a whole bunch of factors just to predict how likely you are to be able to repay a loan. That is why I’m questioning if we have even figured out what a successful ICO is. If there is something that is agreed upon as a success now, we might have something to work with.

Interestingly, there are about 450 IPOs globally per quarter. Which means each year there is about 1200 companies going IPO. Despite this number and the boom of machine learning, analyst ratings are still the best way to value IPOs. I think there are a few reasons for this:

  1. Not every IPO is interesting to a person. It is fairly easy to screen out if an IPO is (or is not) in your range of interest.
  2. Most of the time privileged information is still the best way to value an IPO (and I suspect and ICO actually). What is written in the SEC filing (or whitepaper) is just the tip of the ice berg
  3. Success means different things for different companies

Now, I write all this not say the idea is terrible and that I don’t believe in the technology. I think you should just go about this in another way. If this is what you want to achieve eventually, perhaps you should invest your effort in building a platform where you try and establish some factors first then get people to agree on the factors. Then get people to score the ICOs. With that you can collect data. Then you can get start suggesting what your AI actually thinks is a score. Machine learning is great as helping us evaluate things but not great at helping us invent a way to evaluate things.


Thanks Rigardio. Will put this in mind.

We are currently working on our Trufield project that will provide free electronic medical records systems in poor developing countries to help improve healthcare. It is a great humanitarian project that will revolutionize healthcare for millions of people. Further information here at http://Trufield.com.