Yomi.ai - The love for reading

[mks]

It has been a long time coming; today, I finally launched Yomi.ai for the public.

Encouraged by our 2018 survey and the surprising choice by the great crabigator to pick our No Fluff: Katakana app for the best tools for January 2019, I decided to build a useful resource for Japanese-language learners who aim to improve their reading skills.

I have built some prototypes over the years for learning apps and websites, showcased it to teachers, students, angel investors, and other business people. Side-note: The business people I spoke to urged me to go after Chinese-learners. While the sentiment was generally positive, nothing really took off until earlier this year. I had pre-selected a group of people from our survey (participants who picked a growth mindset, are fluent or literate in English, are most interested in learning Japanese, prefer a website, would pay for a service with the features they want) and then picked participants randomly from that group to collect continued feedback on our development and built what you can see now.

The goal is to become the number one resource for your reading needs (and mine to pass the JLPT N1 reading section).

While what you see (I’m a little bit worried about our database crashing when I open it to all of you at the same time… I’ll keep an eye on the logs) is just the beginning of this endeavor, I refrained from using terms like “open beta” or versioning the service. For me, it takes away from the responsibility to rise to the self-imposed challenge and it gives permission to not work for the community of language-learners that I am so grateful for.

At the same time, I also want you to know that we at Yomi.ai acknowledge that there is much to do. I hope you are excited to be part of the journey.

I’d like to read your thoughts, constructive criticism, encouragement, questions and other comments of that nature.

P.S. If you took part in the survey, again, thank you so much! If you also left your email but haven’t been part of the pre-launch development phase, look out for an email from team@yomi.ai over the next few days. We’ll be drawing the winner of the $20 gift card later today.

---

Edit:

Q: Can I import burned Kanji and Vocab from WaniKani?
Yes.

WaniKani kanji import now live on Yomi.ai

Q: What else do you have?
Besides the website features, we run different courses (Genki I course and a Manga Reading club at the moment), and communicate with each other also through discord.

[TrinityBringer]

Sounds ambitious! I’d love to see it in a more finished state, especially with the grammar explanations.
Will this service run on donations, or will it be paid one?

[mks]

Thank you for checking this out.

I would love to be able to keep Yomi.ai running on donations for a while.
With that being said, I think we will be in a position where we will have to turn the switch and charge for premium features that are more expensive for us (infrastructure and development / personnel costs primarily). By the time we can create recommendations based on your wanikani and other vocabulary data, this is most likely to be the case. I recognize that it’s tough to charge for something that has been available for free previously, but we also did not want to have a lifetime sale without a substantive amount of our first feature set built out.
We were deciding for the supporter model, because it seemed to be a good win-win solution from a company/customer-perspective. We as a company get to work closely with our user base while they benefit from shaping the product and incentives on reduced fees if and when we switch to a paid service.

We’re also currently exploring to extend our communication channels with our donors while increasing our development speed.

[Kumirei]

Could you talk a bit more about the features? What do I get from this that I don’t already have with Yomichan and other apps?

[trunklayer]

I’m sorry, but for some reason, https://yomi.ai/ doesn’t load for me…

[Kumirei]

Loads fine for me

[trunklayer]

Must some problem on my end then…

[Ninkastmin]

I loved the url name!!!

I am about to fall asleep but I’ll be checking this out in the morning

[I-am-ice-cream]

Ooh, I’ve been meaning to improve my Japanese reading comprehension.

Bookmarked, thank you!

[Radish8]

I’d also be interested to know how it differs from e.g. Manabi Reader, Japanese IO? Is it that the focus is on books / novels, rather than articles and other web content?

Simple piece of first feedback; it’s kind of weird to read something and not know how long it is. Selecting a given work starts you off at the beginning, but gives you no sense of the scale of what you’re reading. I’m also unsure whether if I quit it will remember where I got up to? Particularly as there doesn’t seem to be any navigation by contents page to help me get back to where I was quickly.

[Myria]

I had a look, and I don’t really like the katakana for the furigana, it’s really hard to read Also, since the furigana are randomly generated, they fail in some places, especially since those older works sometimes are heavy on weird place names with weird kanji, or using old versions of kanji, and those sometimes don’t get any furigana at all, which makes it very confusing and sometimes wrong.
  
Also not a fan of the furigana above the okurigana, that’s not the way a book would do it.

I think it’s gonna be quite difficult to get accurate furigana on all the kanji, also given the age of the works you’re covering.

(Also 。 and 、 seem to get mini versions of themselves as furigana)

[HotWeather]

No login page?
Confused

[morteASD]

I read the whole page but still have no idea about really what this thing does. I feel like starting the about page with ‘Mission driven development’ shouldn’t really be priority Like what are the features? What does it do? Who is it for? Everything seems pretty vague to me.

[Naphthalene]

I don’t have time to read the replies right now, but I had a quick look at the 漱石夏目 page, since it was featured, and I’m a bit surprised by the selection. First of all, it’s a lot of prefaces or explanations rather than the book themselves. Second, it only has the preface of the second and third volume of 吾輩は猫である. What about the first volume? Is that normal?

[mks]

I have to apologize. You are at no fault. If you don’t mind, can you share with me (either here or via private message or email to team@yomi.ai) some information that could lead to the resolution of the issue? Browser-information would be a good start.

[trunklayer]

Actually, I’ve just tried it again and now it works!
Whatever change you’ve made must have worked!
Thank you!

So far seems quite interesting. I especially like how text is given one paragraph at a time. It’s very friendly for readers, especially, for beginner readers.

[mks]

Thank you all so much for the feedback. All of your posts have been very helpful.
I’ll respond to individual posts later today.

[mks]

There is no need to collect user information at this point, other than to offer tighter contact channels to drive the development of the site. So there is no login functionality as of now. This will change soon, however, because we want to collect information on what vocabulary you already know. That way, we can surface you recommendations based on that. To do that, we will store user data.

It is also true that competitors usually have a nice landing page where they collect an email address before ever showing anything of their product. We decided against that route.

I hope that makes sense.

[ninacchi]

ありがとうございます

[mks]

Thank you for the feedback. We switched the furigana to hiragana. If there’s people who would actually prefer to have the katakana for furigana, we’ll provide an option to toggle between the options.

I wouldn’t go so far as to say the furigana is randomly generated, but that especially older versions have given us trouble. We will try to create a seamless solution that allows for easy, verifiable correction.
A better okurigana presentation that’s in line with how books do it will be on the agenda. This actually stems from an earlier prototype where we replaced the kanji with the hiragana reading on hover. This worked better for sentence-by-sentence presentation, but here we decided against it. Not a priority for us as of now, but it will be done.

I have actually not seen the mini versions of 。and 、 myself. Where did you see them?