Should I developed it?
I launched an MVP for an app where you can have conversations in any language and in any situation, such as practicing English in a Starbucks or during a job interview.
The idea is quite simple, at least for the first version. The plan is to add more functionalities than just conversation in role-play mode, such as bilingual dialogues where you can converse in your language but receive responses in the language you want to learn, among other features.
So why am I here? Is anyone else interested in this idea?
I developed an MVP to better showcase what's in my mind, and if I see some interest, I'll continue developing the product.
Link to the mvp => https://www.lantalk.xyz/
Trending on Indie Hackers
It's currently not working. In any case, I believe that there are a lot of apps out there doing the same thing. So the question you have to ask yourself is: How do I differentiate myself from the others?
Yep, that's always a very good question!
But I think more than an answer, it can be a path to follow, as the differences will emerge over time. Btw, for learning purposes, I will try to make it as free as possible.
I also have pending to investigate why it's not working. I believe that on some devices, it's not working at all. Thanks for letting me know!
Very interesting idea. You should definitely build it out.
I have a very strong thesis on the language learning industry. As the world becomes more globalised, the need to formally learn foreign languages will increase. And the willingness to pay is high.
I'm currently exploring building language learning tools which can be incorporated into a person's daily life and they can learn subconsciously.
I'd love to try it out. Happy to brainstorm further 1:1 as well :)
lol gonna say you to DM me but just realized there is not DM in indie hackers hehe, just opened this discord server for in deep talks => https://discord.gg/qCQRPHu9
You should.
I'm on the way! Stay tunned for the first beta! I tihnk it will be in a month
Okay. Looking forward to seeing your beta version.
looking ffwd to it
The demo isn't working at the moment, and I was really looking forward to trying it out. As a non-native speaker, I always want to practice my English conversation skills, but it's quite challenging due to the limited number of people available for practice.
What tech stack are you using for this product?
mm I think it wont work in some devices, I'm investigating why, and also making that when the system speaks there is some signal as its hard to realize when system was talking.
I'm on the same situation about my english, need confidence when seapking...
Stack is angular 17, for the demo i've used edge functions, and the AI is mixtral at the moment, tested with gpt 3.5 also but results are pretty similar.
Value proposition is not being clearly defined and delivered. Not quite sure what can I benefit from this, and if I'm not the target audience, then who might it be.
it depends, I'm on the way to define an small target audience than every one that wants to learn english..., but I'm looking the best way to do it, what I'm looking for is people like me that needs to earn confidence while practicing in conversaions in role play mode.., like starbucks or a job interview..
I dont know what are your concers learning other languages.. if you let me know maybe I can say you if the product will be for you or not.
Btw once I've the first version if you have some interest in learning any other language.. just give it a try it will not cost 1k$ ... and for sure I will try to add some free tier as I think it can be a useful tool and I want the people to use and learn.
Keep in mind Samsung S24 now features automatic call translations and Google Translate phone App can be set to conversation mode which seems to be similar to what you're describing?
well its not exact a call transaltion, is just for a "solo player" user as you will have conversation like if you are in starbucks but with the system, so you can practice any language and in any situation, like you need to go to the airport but you dont feel confident in X language, then practice the situation you need before do it..
But I can do this with ChatGPT, it can also engage in a multi-language conversation.
I can do a prompt "Lets simulate I'm going to the airport and everything is in language X but I only speak language Y, can you step by step give me an overview of what phrases and signs I should be able to read and their respective translation?"
Super easy, I can even have ChatGPT ask me questions and engage in full conversations on a foreign language, with translations included.
Sure, you can access it by paying $20 monthly, but the idea is to go beyond just prompts. We'll track your improvement, offer more modalities than just role-play scenarios, provide instant feedback, and guide you through the learning process. We aim to offer more than just conversation practice; for example, if you're unsure what to say, the app will provide helpful suggestions based on your proficiency level.
The main concern is whether we can leverage AI from the "OpenAI" suite, but there are many other possibilities using its tools.
And btw that's a good point, as I believe it could be the "common" question about AI products these days.
ChatGPT is free, you don't have to pay $20/month.
lol thinking the audio thing is only for chat gpt plus, btw, the above response include many other things that I think can be most useful that only chat gpt
I was referring to ChatGPT as a way to learn in text, not with audio.
For audio you have Google Translate or you can grab ChatGPT text and paste it into Google Translate to have it speak it.
Ah, so yes, you can do those kinds of things, but you can do almost anything with everything by yousefl is just time.
At least, what I'm developing will help to reduce friction in setting up everything customized for your studying improvements, making it more personalized and helpful for you as a user.
Btw, very good points.
yep these would usually be objections from potential users, in the end the USP would be around learning being easier and quicker and ideally some kind of real-time assistance when they're in the situations they need help with
Completely, the USP will be the "learning" more than the AI, as, like you said, everything through AI can be almost done with ChatGPT already.
Can't you do this already in chatgpt with the voice feature?
yes and no, as I think every AI product can be done by some LLM (chat gpt), but the idea behind this one its to be your english mate that you can speak with them whenever you want and test diferents situations while progressing,.
Can you speak with chatgpt with voice like this? Yes, but you will need a lot of prompts..
Can chat gpt track your learnins and auto improve the language level? NO
I see now. I think that it can be useful especially with the "auto-improve" feature that you mentioned.
Yes, I believe it can be very helpful for many people who are frustrated with learning any language (like myself, a little bit with English) to improve step by step with some guidance.
Hi Mark, I tested your demo, but when trying to speak, nothing occurs. The microphone button is not prompting for mic access.
What inspired you to build it?
Hey I'll take a look btw you have to wait that the chat speaks its answer (first time it takes 4s more or less on say the first sentence and if you have your device silenced you will not notice) I'll add some signal when the system was speaking to can identify better when to speak and when to listen.
Get inspired a little bit from myself as I always want to be fluently in english but not many situations where to speak.. before this I developed an other app to make phone calls (another history) and I was thinking why not use the same system to practice conversations...so tried to validate the idea a little bit throught reddit and looks there is some interest and a lack of tools like that, and now here Im I will procede to make some role play conversations with better voice and improved conversations system and more conversation situations.
And if u still have problems with the demo let me know please, just activate the continuous speaking and it have to activate by itself ones the system stop speaking.
Who is the target audience? And what type of people will get the most value from this.
I'm studying who can benefit the most from using this app. Targeting everyone learning any language is too broad; it's like throwing rocks into a river to cross it.
Currently, I'm investigating which "big" groups of different language learners there are, such as those studying for travel or interviews. After grouping the different people studying English, I'll delve deeper to segment them further.
So dont have the right answer yet..
The demo is a bit annoying, I tried speaking in two languages and it keeps misspelling. On the other hand, it switches between languages once, but not switching back to first language. Lastly, I'm not a huge fan of stock photos like the ones you added, they can be found on many sites, you have to stand out. Keep going!
Yep, I'll have to improve the demo. At the moment, it's only accepting English, so I'll definitely have to inform users about that.
I'll also take note of the feedback about photos!
Thanks for the insights.
You lost me when I have to give my email for a demo. People would not want to give their emails this early, they dont want to get spammed when they are not even sure what or how good the product is.
Do it after they have actually tested the product and have there aha moment.
-----------
Anyway I think this would work much better as a paid not premium b2b tool.
Unless you are a language influencer. Or would partner with one.
------
I think this could work, if you are willing to do the work to acquire users. This is more of a marketing problem than a engineering one
----
Try focusing on more high value use case with people that have money for this.
People who are looking for jobs are typically unemployed and don't have much disposable income for this. Its a lot harder sell.
----
Don't build more until you out reach to people that would know more of the problem.
About email completly agree, I removed it today morning as I see many friction for testing.
Not planned about being not premium b2b tool, but can be a good option the only question I've who can be the target of this kind of tool in the b2b world. Maybe companies that give as a perk to employees? I dunno but I'll investigate for sure.
Yep yesterday someone recomended me this article on YC https://paulgraham.com/ds.html and as the almost first sentence says:
'The most common unscalable thing founders have to do at the start is to recruit users manually. Nearly all startups have to. You can't wait for users to come to you. You have to go out and get them.'
So completly agree that there will be hard work on marketing as I dont have much money too...
Sure, but to add high value I'll need to develop more.. That I think there are on my plans now on start developing the product, at least an "extended" mvp where you can login and feel with a better experience...
I think here not agree at all hehe, I'll build a little bit more as I think I need to validate more things and need a little bit more of development, btw I need first too study who will be my market fit audience.
And many thanks for all the info!
your welcome
I recommend b2b cause b2c is shitty. you can make it work but it will take longer and distribution is the main problem.
For b2b applications. Here are some ideas.
b2c can work too either you build an audience or partner with people that have an audience
your app as it is, I think would work better as a mobile app. thought you have to test it your self and distribution is still a problem.
You have sweat equity.
Dude a login isn't a core feature. It does not solve a user's main problem. Its something they have to do to use and secure a app.
Having a login isn't going to magical make your app better.
Your problem is that you have no idea what feature to build that is actually valuable. That's why I urge to you to talk to people that teach speaking to see what they actually need.
---------
I recommend REWORK to know how to approuch building stuff.
And TK Kader's yt channel for the strategy. He's a founder that exited
Thanks again for you interesting response!
About going b2c or b2b as I'm not much experienced on sales or marketing (my background is engineering) I feel much harder go b2b than b2c (for sure my own prespective), as I dont know the tools on how to search companies or how to sell the product to companies btw can be interesting to read and learn about it, and compared to b2c that for sure is harder than b2b I feel that is like more searching for sites where people can be interested about the product, like english learning communities (or any languge), travelers communities that wants to know to minimum for comunicate when traveling to other site... I moved a little bit on reddit and asked some question about if the product can be of interest etc... But what I mean at the end is b2c is easiest right now than b2b for me.
About login what I mean is not login perse, is to add more features but for sure I'll need user auth to do things like continue conversations, add more value by having a little bit more of data about each user, role play conversation in diferent scenearions user wants...
And the partnership with language influencers is a good idea too, like some referrer system or similar system..
A video or audio sample on the landing page would make it more interesting
It's not clear what do I put into the system vs what's just random, how much guilded would it be
I'll take note, and yep video can be a good one to show the product too!
Thanks!