We bet you came here with a particular question in mind and thought of reading through multi-paged user manuals let’s say does not excite you. We get it, we think that way too.

So, here are your answers and if you still cannot find the one you came for – we are all ears. Ask us!

YasNah provides a completely intuitive and familiar cross-platform experience for textual data labeling activity. With a simple swipe on a smartphone or tablet screen, click, or key-press on a desktop a labeler puts their "True" or "False" label.

For a dataset creator, YasNah provides interface and workflow to upload the initial unlabeled or partially labeled data, assign members and manage access to this data and retrieve labeled results.

YasNah is not an automated data labeling solution. It requires the human labelers workforce to label the data. There is no algorithmic or machine-generated assistance provided for labelers. Previously assigned labels are also not visible to the current labeler to avoid biasing.

YasNah does not provide dataset creators with a workforce for labeling. One should recruit and invite labelers for their data on their own.

YasNah does not support the labeling of any other data types than texts.
Images, video, audio, sensor data is not supported at this time.

YasNah is not a marketplace so far and there are no internal mechanisms for backward value flow from dataset creators to labelers. If there is a need or desire to incentivize and reward labelers, then it must be done using means outside of YasNah service.

YasNah is not an annotation tool. There are no features to select spans of text and annotate them either with comments or with categories. One would need to split the textual data into such spans and formulate general (Yes/No or True/False type) questions to be asked.

Similarly, YasNah is not optimized for multi-class categorization labeling. One could "flatten" the data in a way that each example appears with multiple general questions asked - one question per class. E.g. if you have 100 examples to be categorized into 5 classes that would require 500 questions to be asked and answered in general Yes/No form. Possible, but not optimal, we know.

Yes. This is one of our main advantages. We built YasNah with the convenience of the user in mind.

Our app for iPhone and iPad can be found in App Store℠ here:

Our app for Android™ can be found on the Google Play here:

Get it on Google Play

We also provide a web-app to use YasNah on any desktop computer, which is also is responsive and optimized for mobile. Access it here:

For now, we only support texts. Ideally, those would be relatively short snippets of text ( one or several sentences, a paragraph or two at most), so a labeler could easily and quickly ( on average should not take more than 1 minute per snippet ) read it from a screen as small as a smartphone's.

We do not support other media data types like image, audio, video, or sensor data labeling at this time. It will likely require us to create separate apps with labeling experience optimized for each type of content.

You are welcome to post your need or idea on our Ideas portal here:
We would love to hear from you!

At this time YasNah is not providing a workforce for data labeling. Whoever needs their data labeled should promote their cause, recruit and invite their own labelers outside of YasNah.

However, you may decide to publish your project (we call them Topics). This way anyone accessing YasNah will be able to volunteer and join the labelers workforce for your cause. Note that you will not be notified or have any power to moderate who joins, labels data, or leaves your public topic. You will still be able to manage who has administrative access and delete unwanted members. But nothing will prevent them from re-joining again until you remove your topic from public access.

We provide access to our backend API if you are in need to customize labelers' experience or automate some processes.

You can also feed data in and out through integrations API from your own servers or gateways with third-party providers.

Documentation for our integrations API is available here:

If you have further needs and do not know how to implement a solution for those we are ready to talk and see how we could help. Tell us what you need.

YasNah is simple. Have you seen a simple data labeling platform before? We would love to hear from you about it then. Because we needed one and could not find it in 2020. That's why this product was born.
Simple task, simple process, simple outcome, and simple pricing. It takes a lot of work to make things look and feel simple and we are here to do that work for you.

YasNah is focused. Even market leaders do suffer from feature creep. Trying to do too much in a too small of a space of a single app leads to subpar experiences and loooooong learning curves.
We would love to cover more types of data and machine learning problems to solve. It just does not seem to play well when one tries to build a one-size-fits-all tool. We chose to focus on one particular type of task at a time and make it the best experience we can. Quality and simplicity above quantity and complexity remain our goals.

YasNah is mobile. Most of the tools on the market are enterprise targeted software available only on desktop. You know how it goes regarding UX in business - who cares that it sucks for those who directly use it when it solves the problem for their superiors and managers.
We built YasNah to be a mobile-first and user-first experience for labelers. And we did not forget about dataset creators' administrative needs either. We built this tool for the user's happiness, productivity, and convenience first, no matter who pays the bill.

YasNah is unlimited. Other similar tools model their pricing on artificially imposed limits like the number of people in the team, or a number of annotations/labels that can be done per month or sometimes even on the quality of the resulting dataset. Yikes!
We do not impose any quotas or limits on what you can use YasNah for. Need a million examples labeled - you're welcome. Need a hundred or thousand labelers involved - bring'em in!  There are no limits on the number of questions, labels, or people involved as labelers. You get it all at once for a flat unified subscription price.

YasNah is always free for labeling activity.

If you need your data to be labeled we offer you a 15-day trial to see if things work great for you. If you manage to achieve your goal in that period of time - amazing! That is what we want too! To make it so fast and simple that you do not even need to pay.

If your project is bigger than that - YasNah has the simplest pricing model possible. $10/month per creator account. Remember - you only need 1 account to put your data out there and get the results in. Labeling activity is and will always be free!

No, not by YasNah at least. For now, we are not facilitating any kind of value flow from creators to labelers either.

If you want to be incentivized or rewarded by a creator for your work - you need to contact them directly outside of YasNah.

We might consider implementing this in the near future, though - let us know how much you would like that on our Ideas portal:

The funny part is that we technically do not have a user "accounts" system in its classical meaning. There is no particular table in our databases with logins, emails, and passwords (hackers can sleep well without bothering). We encourage you to use popular credential providers like Google or Apple® and login through their authentication system. This way your passwords and other personal data are managed by those giants and not by us. Recovery in case of lost or forgotten credentials - same.

We do not have accounts because we do not need to have them. Simple, right? When you subscribe with us it is our payment solutions partner Stripe who gathers and securely stores your billing information and subscription status. When you create a topic inside of YasNah we record who created and who manages that data object as part of it. Your email is an attribute of your data in YasNah, not the other way around.

So what that 'Sign in with email' option is about? Continue reading the answer to this question right below.

Currently, that is an internal feature for testing purposes. There is no way to create credentials for that as those are hardcoded (hashed and salted of course) deep inside our backend servers.

We keep it for things like apps review by Apple and Google and just in case if something will go awry with their sign-in services (which is extremely unlikely, but black swans do exist, you know).

The best option - to our app on  Web    iOS™    Android™

A subscription also does not sound like a bad move - Subscribe!

Ideas and opinions are always welcome on Ideas Portal

Detailed features documentation with pictures, yay! will be available here.

 Apple, the Apple logo, iPhone, and iPad are trademarks of Apple Inc., registered in the U.S. and other countries. App Store is the service mark of Apple Inc. registered in the U.S. and other countries.
IOS is a trademark or registered trademark of Cisco in the U.S. and other countries and is used under license.
Android, Google Play, and the Google Play logo are trademarks of Google LLC.