Launch YC S21: Meet the Batch, Thread #7 | Hacker News Reader

Launch YC S21: Meet the Batch, Thread #7

91 pointsdang135 comments

Here's the latest Meet the Batch thread for YC's S21 batch. The previous thread was https://news.ycombinator.com/item?id=28156460. For the original description see https://news.ycombinator.com/item?id=27877280.

There are 5 startups in this thread. The initial order is random:

Monto (YC S21): Salary on Demand for Latam - https://news.ycombinator.com/item?id=28209932

Plug (YC S21) - API for multiple payment providers - https://news.ycombinator.com/item?id=28209933

Abstra (YC S21): Figma-like front end development for web apps - https://news.ycombinator.com/item?id=28209934

Levo (YC S21) - High yield savings account for Mexico - https://news.ycombinator.com/item?id=28209929

BoldVoice (YC S21) - Accent coaching app for non-native English speakers - https://news.ycombinator.com/item?id=28209930

View on HN

We’re Anada and Ilya, the co-founders of BoldVoice (https://www.boldvoice.com/). BoldVoice is an accent coaching app that helps non-native speakers learn and practice the standard American accent.

As immigrants to the US, we’ve personally experienced that having a foreign accent can affect your confidence and job opportunities (by as much as 20%, according to studies). So we built this app to help non-native English speakers like us.

Users start by watching 3-5 minute video lessons to learn new concepts from Hollywood speech coaches, like Ron Carlos, such as how to pronounce tricky sounds, or how to properly use intonation. They then spend around 10 minutes a day improving their accent skills by going through words, sentences, and conversations, with real-time feedback on their accent from the pronunciation assessment AI model.

The user tells us what their native language is during onboarding, and we serve them content on the typical accent and speech challenges that a person from that linguistic background faces in English. Regarding the feedback part, the user speaks certain words and sentences into the app, we record it, we compare each sound to the "correct" sound according to our library, and we spit out a score and a recommendation for what they should change.

The vast majority of language learning apps focus on grammar and vocabulary, not on speaking. Even when they do speaking, they score it broadly as "right" or "wrong" with no feedback on how you can improve. Learning an accent is all about muscle memory formed in the mouth/lips/teeth, not memorization in the brain—this is what most language learning apps get wrong.

Practicing the physical skills required to acquire an accent is a lot like working out, and with consistent work come results -- usually you can start seeing a difference within 2 weeks. Since launching 8 weeks ago, we’ve had startup founders, globally distributed teams, BPO firms and even middle school Spanish teachers pick up BoldVoice to start working on their accents and feel more confident in English.

We'd love to hear your experiences with accents and your comments about BoldVoice!

Loving what you’re doing here - excited to show this to my non-native speaker SO and give the AI assessments a test drive.

Thanks! Which language is your SO's native language?

Feel free to send any good (or bad) feedback to founders@boldvoice.com!

Japanese :) first gotta pitch her on it though lol.

We're excited to help and be on standby :) Feel free to mention that BoldVoice is made by non-native English speakers for non-native English speakers.

Our dialect coach Eliza Simpson also recorded an introductory video that lays out the values & mission very well (2:58 mins): https://vimeo.com/572809215

How do you decide what the 'standard American accent' is?

I spent some time on a remote team working with a great group of folks from the US South. I'm Canadian. I picked up a lot of accent changes that stick with me today ("y'all" is the best plural second-person pronoun ever).

But even things like California vs New York, there are differences. Urban vs rural. Minnesota vs Florida.

There are interesting implications in how you make those choices. ie: "Well, however I speak is correct" will get you elitism accusations I'm sure.

We can go down a very deep rabbit hole with this question, because you're absolutely right there isn't just one accent spoken in America.

But for what it's worth the term is defined as the "umbrella accent of American English spoken by a majority of Americans and widely perceived, among Americans, as lacking any distinctly regional, ethnic, or socioeconomic characteristics". You can read about that more here: https://en.wikipedia.org/wiki/General_American_English

If you do have the chance to jump into the app and watch some of the videos that our speech coaches, Ron and Eliza, have produced, you may notice they also refer to their own accents as "this accent" as a nod to the fact that ultimately what they are able to teach is the accent they are themselves demonstrating.

Pure GAE/SAE is fairly arch and stilted - I assume you are chilling out on some of the more arcane aspects like liquid Us in "reduce" and "student" and "forehead" as "fahred"? Most dialect coaches default to something looser when teaching an american accent because SAE sounds almost british if you're strictly adhering to it.

If by "liquid U" you mean pronouncing /ju/, then no, GAE pronounces it "redooce" and "stoodent", not "redyooce" and "styoodent." From the article:

> yod-dropping after alveolar consonants (with new pronounced /nu/, not /nju/)

That's interesting - Speak with Distinction, which was treated like a bible of SAE when I was in school, leans on /ju/ for stressed syllables:

https://books.google.com/books?id=hM-GDwAAQBAJ&pg=PT126&lpg=...

not my area of expertise, but I had assumed from the article that SAE and GAE are equivalent, but maybe not?

Hey guys, for the record, in our description we used "Standard American accent" as a bit of a catch-all term, not quite in the most technical manner.

One of our dialect coaches, Ron Carlos, slacked me the following blurb: The accent taught in American theater schools (which we're calling SAE here, most would call it Mid-Atlantic) is an accent that was popular in Hollywood in the 40s. Many schools have modified it to sound more modern. And we definitely don’t use that accent in BoldVoice. Ours is more modern: for example, can’t and Cod will have two separate vowel sounds, and Tune and toon will sound the same.

Okay, that's helpful, thanks for going right to the source - yes, that's what I learned in theater school (referred to as SAE), and found that it was useless for everything except Shakespeare and poetry. I thought it pretty unlikely that was what you were going to teach non-native english speakers.

I'm guessing that the meaning of "General American English" has drifted over the years - it seems like Speak With Distinction is from 1942, when those sorts of pronunciations were common, and it would have been appropriate to teach that to non-native speakers. Now the accent is a little bit different, just like how "business casual" now includes certain types of jeans.

Having an accent increased my chances of success in US. I had an Italian strong accent. When my English got better, people thought I was from Mexico. People were not as fascinated to the accent and lost some interest. I think the problem is not necessarily the accent, but how open people in front of you are to new accents and how confident you are with yourself being a foreigner. I'm interested on how the product will develop!

Thanks for sharing your story! My cofounder Anada and I can both relate, as we're both immigrants to the US as well.

Something we want to stress with this product is that we want BoldVoice to live up to it's name -- helping non native English speakers feel bolder and more comfortable when speaking English and be more clearly understood.

Pretty cool product out there ! I come from the Conversational AI background ! Can imagine the hard work you've done to get this right on the accuracy ! Looking forward to testing it out ! Wishing all the best !

Thanks, appreciate it! If you have any feedback or thoughts to the pronunciation analysis feel free to send them to founders@boldvoice.com!

I disagree fundamentally with the premise here. An accent is simply someone applying the rules of pronunciation from one language to another. No accent is right or wrong but your product seems to imply that North American accents are desirable or correct while accents rooted in non-english languages are less desirable.

Instead of trying to "acquire an accent" why not focus on clarity of speech and pronunciation? If someone is unable to understand an accent, it is more often a case of that person not being exposed to differences in pronunciation which speaks more to the listener's lack of exposure than something wrong with the speaker.

Many regional accents from the UK are difficult to understand because they use non standard grammar, sentence structure and place stress and intonation differently.

I agree with the focus of most existing language learning - vocabulary and grammar are the foundation. Intonation and pronunciation are important. An accent is just top dressing - it is relatively unimportant in the grand scheme of clear communication. The goal should be clear communication, not a north american accent.

Thanks for sharing your thoughts. I invite you to watch one of our dialect coaches, Ron Carlos, share a bit about both the values and process that he brings to the table at BoldVoice (1:43 mins) - https://vimeo.com/572809188

As our name implies, we want to help non-native English speakers feel bolder and more confident when speaking English. To feel more confident, indeed, one may need to more deeply learn the rules of English grammar, and more varied vocabulary.

However, in the pursuit of more confident and clear speech, one may also want to learn how to adapt their pronunciation, how they place stress, how they use intonation and they way they use pacing and rhythm. All of these are the component parts of an accent.

A note from Ron (who just Slacked me while I was writing this): Clarity of speech and pronunciation are big parts of an accent. You can’t change those without changing an accent. We are dialect coaches. We love peoples’ native accents, which is why we use sounds and samples from each speaker’s familiar language to teach sounds in American English. We want to empower our users to be able to control their speech so they can choose when sound more American in situations where it may benefit them. We truly hope that one day accents won’t matter, but until then, we have folks who feel embarrassed about their accent which keeps them from showing up with their full selves. We’re here to help those folks feel more confident with their speech.

You have not addressed the fundamental point of my comment - why are you selling a north american accent in particular?

Clear speech is not predicated on acquiring a certain accent and i feel you are feeding into the stereotype of an accent determining the quality of the speaker, not the content of speech.

Edit - you ignore the fact that a native speaker of English may have an accent that is not North American. Think about millions in the Indian subcontinent or Africa. Are you not being subtly racist in implying that only accents from North America are desirable? Personally, i find many African accents incredibly clear and easy to understand.

This is the reason for my opposition to "accent coaches" - they focus on the wrong thing - the accent. To anyone with half a brain, D Trump sounds incredibly stupid despite his north American accent.

Many people learning foreign languages have a goal of sounding like a native speaker from a particular area. There is nothing wrong with that.

As a fellow immigrant working in the US and as someone working with non native speakers from other regions in the world, I can assure you this is not about racism. It is about team cohesion, clarity, getting things done in an international team. People have to spend extra brain cycles just to tune in to how you say things, before they can focus on what you actually say. This is especially true at the beginning of working with someone who's accent you have never heard. There is a big difference in accents between someone from India, France, Ukraine. While there is nothing wrong with that, it has value for international teams to mitigate this difference and settle on a common standard.

I like your point about cohesion. I'm not an immigrant here so, perhaps obviously, I agree. I have worked on teams and accents of those I have to collaborate with is a big factor in determining overall "friction". Often if the accent is too strong, it really makes the interaction dreadful because I (and others) want to understand you but it's difficult and embarrassing having to ask "what?" three times per sentence.

I don't think you need any better example than support call centers. How many people routinely avoid calling for support or simply loathe the idea of doing so because they're like to get someone with a heavy Indian, Filipino, etc. accent that leads to the scenario I mentioned above ("whats?")? When it comes down to it, it has a real cost in many ways.

Sorry for veering offtopic, but can you please stop posting in the flamewar style to HN? You've unfortunately been doing it a lot, in many threads. It degrades discussion and, as https://news.ycombinator.com/newsguidelines.html explains, we're trying for something different than that here. Note this one, for example: "Have curious conversation; don't cross-examine."

The main thing that would fix this is if you edit the swipes out of your comments, e.g. commenting on how bad the other person's point is, how they are nitpicking, how they are failing to argue properly. If you simply make your substantive points directly, without negative 'you' statements, your comments will be much better.

Also, please omit flamebait like the bit at the end of your comment here. It breaks the site guidelines too, and leads nowhere good.

Disagree with you here. For some people, accents can be a big personal issue. This helps solve that.

The NA English accent is considered the most clear and easy to understand. This is not a controversial opinion and is held by virtually all non-native English speakers around the world.

Frankly there are billions of people who would kill to speak English like a native American. It’s about economics, not identity politics.

>The NA English accent is considered the most clear and easy to understand.

Do you have any evidence of this? I would believe that it depends strongly on the language(s) you already know. Perhaps you are confusing prevalence with clarity. North America has tremendous soft power - more so than any other part of the world and their accents are ubiquitous.

Clarity is not necessarily function of the accent.

American media (movies, news, etc.) is universal. People all around the world watch American shows and movies and are accustomed to the accent. This output dwarfs any other versions of English. So, to begin with, people are just more familiar with American English.

Adding to that, I’m not confusing prevalence with clarity. The standard American accent is clearer and easier to understand than most other variants, including American subcultural accents like the Boston, Texan, or Southern accents.

So you have no evidence of your claim.

>, I’m not confusing prevalence with clarity

...but you do that right here:

>American media (movies, news, etc.) is universal. People all around the world watch American shows and movies and are accustomed to the accent.

So it is prevalence and not clarity.

No, it is both, as I just said.

It is widespread common knowledge that the American accent is the easiest to understand. If you don’t agree with this, I’m sorry but you are not speaking from experience.

Can you identify any intrinsic property of General American other than media, business and language school familiarity that makes it particularly intelligible? Its pronunciation is as divorced from phonetics as most English dialects, it elides some consonants and vowel clusters that are distinct in most other English variants and it tends to be spoken more rapidly than some English dialects.

And yes, I have considerable experience of speaking English abroad, and whether I think the listener will understand me better if I exaggerate my British diction, adopt more General American-sounding syllables, pick up quirks of pronunciation and phrasing used in the local ESL lingua franca or pronounce a particular problem word as it is written is entirely situational, and generally based on what the person I'm speaking to is most exposed to.

> It is widespread common knowledge that the American accent is the easiest to understand.

I didn't know that this was established. I'd be curious to see a source for this — has it been studied?

As a French, the Amerian accent is usually the easiest to understand. That's also what most people around me think, though they also think of the British accent as more "fancy". This is for the "American you hear on TV" accent. which is also the one that I hear the most on Youtube or sites like that.

You're right about clarity not being a function of accent, at least not totally. There are some people that speak terrible English with a French accent, and others that speak clear and easy to understand English with a French accent. The same applies to pretty much every accent. I also agree with you on the soft power, however part of the soft power will mean that people are more used to the NA accent.

As a final point, I've heard multiple times that sometimes people that speak English as a second language can understand each other really well while Americans have a really hard time understanding them. If that's a real phenomenon, then having a non-American accent could be detrimental if you want to work in the USA.

American English is easier to understand because it's less distinctive, it has lost any character, which is common for linguæ francæ.

So I don't think it's solely about American soft power, although that will undoubtedly play a part.

> Frankly there are billions of people who would kill to speak English like a native American

Probably, at least in poor and undeveloped countries. But a lot of people would kill even more to speak with a posh British Accent, it's really hard to sound sophisticated when speaking American English.

Unfortunately some people do think differently of you based on your accent, subconsciously or consciously

Not to stir the pot but: Wasn’t this controversially brought up as a signal among startup founders by PG?

"The goal should be clear communication, not a north american accent"

Yes but how do you teach clear communication without focussing on accent ? You ultimately have to pick a way to speak and accents define that way.

This hasn't been my experience. I worked with a man years ago that was relatively fluent in English, but whose thick accent made understanding him somewhat difficult. Listening intently was required to understand him and he would often need to repeat himself, even though he was using the correct words and grammar. It's something he ran into pretty often at the time and he worked to improve it. Something like this app may have helped him. Slight accents may be "just top dressing," but this is most definitely not always the case.

> An accent is simply someone applying the rules of pronunciation from one language to another.

Exactly, and that is the wrong way to speak the "another" language. The ears of native speakers are tuned to a certain pronunciation, any departures from that will be harder to understand, it will be a distraction.

You might not like it, but that's how it is.

> Many regional accents from the UK are difficult to understand because they use non standard grammar, sentence structure and place stress and intonation differently.

True. It's also true that many regional or working class accents in the UK will impair your job prospects in the UK. It's probably better to come with a Swiss or Norwegian accent than a Geordie one, when applying for a job at Goldman.

> The goal should be clear communication, not a north american accent.

I agree with this 100% - mutual understanding is the basis of communication. Nevertheless, I don't think this the problem that BoldVoice is trying to solve here. While I'd love all my audiobooks to be read with a Glaswegian accent, North American one is what most people are internationally most familiar with. As a non-native speaker, being able to speak fluently with an accent that is highly desirable (at least within this century) carries not only practical utility but also social status if you consider the fact that this is someone's second (or nth) language.

This is a native english speaker: https://youtu.be/pit0OkNp7s8

We are a diverse bunch

Wow! That was amazing, I couldn't understand more than a few words of the first speaker. The others were much more intelligible, but it required a lot of attention (which is also interesting, how much of listening is really passive). I feel for any non-native English speakers used to a different accent trying to understand that.

My wife moved from Finland, and ended up working in A&E in Dundee. She had a rough time at first understanding some of the heavier Scottish accents.

I have tested her now, on Rab C Nesbitt, and she passes:

https://www.youtube.com/watch?v=3qeWbC7XtO0

https://www.youtube.com/watch?v=4NriDTxseog

BoldVoice, I hope you get Amy Walker on as a consultant...!

Thanks for sharing!

That sounds like a very nice idea! I'm French and not confident at all in my oral English because of my pronunciation. I did a bit of job hunting recently, and I was afraid that the topic of English would come up precisely because of that.

Hey, that's exactly why we're here. Happy to have you check it out and let us know if this is the type of program you would want to keep doing & practicing. If you have any feedback, questions or issues, please reach out to us at founders@boldvoice.com!

Hey we are in the same boat, like many French people I guess! I’m confident speaking English but it feels so frustrating sometimes to not be understood by my interlocutors. Especially by other non native English speakers who usually have a hard time understanding my accent.

I’m definitely going to check this product out, I always wanted something like it.

We're here for you! If you have any feedback, questions or issues, please reach out to us at founders@boldvoice.com! We're constantly working on improving the app and adding more engaging content.

Can you please do the Kiwi accent instead?

Thanks for the suggestion! We'd love to explore how we can best create curriculums for other accents of English, such as British, Kiwi, Aussie, etc. But, as you can imagine, something like this takes a lot of time & planning, so we're being careful not to over extend ourselves.

Interestingly enough, if you do engage with the content in BoldVoice, you will learn the physical skills that help you understand how different sounds are made and how to use your articulators. Once you are fairly confident in these new skills, you may even be able to apply them to quickly learning all kinds of new accents.

Hi, congrats on your launch.

This app would have been ideal for me; I'm now looking to improve my pronunciation after working on my grammar for a while. But I lost trust in the first 15 minutes; if you check lesson 1/4 of "Unvoiced and Voiced consonants," the app expects me to repeat the following phrase with an American accent:

> This house is a part of a low-income housing project.

But the coach audio clip does not say that; the coach says:

> This house is part of a low-income housing project.

So the audio clip of the coach features what seems to be correct grammar, the engine that corrects my pronunciation does not. If I try to say the phrase the way the app describes it, with "is a part", I'm able to pass the exercise. If I say the way the coach says it ("is part"), then it fails me.

So is my perception that the app does not follow American grammar. It is not great to have an app that helps me improve my pronunciation while it screws my grammar.

Also, how do I delete my account?

Hey there, thanks for checking us out!

The error that you pointed out has been fixed, and now the sentence matches the coach's sample -- feel free to double check this on the app. Both "part of" and "a part of" are grammatically correct. That said, we definitely agree that the coach sample should match the written sentence.

By the way, potential issues such as these is exactly why we added the flag feature on the practice cards, through which you can report any errors in spelling or other types of issues. We strive for 100% accuracy, but there are still real people behind this making all the content, so we appreciate user input to make things always better.

If you would like to delete your account, we can help. Just email us at support@boldvoice.com with your account details.

Have you considered adapting your app for deaf people?

A friend of mine developed a pronunciation practice app for deaf speakers of Korean as a student, which won them an innovation award. But according to my friend, the pronunciation score (which was based on comparing spectrograms or something) didn't work all that well, and the award was mostly due to the non-technical team members making up a heart-warming story to promote their project.

I assume your feedback system is more advanced (can you share some details on how you determine what recommendations to make?) and would also work for deaf speakers. But you can't really assume familiarity with another spoken language, so you'd probably need explanations specifically tailored to deaf people.

That's an interesting use case that we could potentially explore down the line, thanks for sharing!

Right now, we're focusing on non-native English speakers, as it's a problem that we personally understand best -- as well as a very large market.

I just want to note that for many deaf people (including practically all people that were born deaf) sign language is their first language, and thus, technically, English isn't their native language.

edit: I'm also curious to know more about what kinds of analysis/algorithms you do to detect the accent, if you can share a little bit. Can your app distinguish between different American accents? Or could it hear the user talking in African-American English and assume it is "incorrect"?

The problem with adapting this to other forms of atypical speech is that their recommendation system likely relies on a catalog the phonemes L2 speakers have issues with (the example most people know is the Japanese "L" "R" swap) so it's much easier to create courses with specific focuses and solutions.

If Google's Project Euphonia [0] is actually still ongoing and they release their dataset/methodology of training models with that sparse dataset I can see your idea as approachable; even accented speech is a tough problem to work on considering how many variants exist worldwide (but their approach looks good!).

[0] https://sites.research.google/euphonia/about/

I tried to build something like this a couple of years ago, just by using speech-to-text, and that worked really badly (but it was better than nothing). I've given BoldVoice a spin and am very impressed by how well it works on phoneme level - well done, congrats on launching!

Thanks for trying it out, glad to hear you liked the tech!

We're continuing to improve and expand both the technical capabilities of the pronunciation assessment, as well as add more varied & exciting content that a user can go through.

I have very mixed feelings on this.

First of all, it feels like a patch for latent discrimination. I'd rather strive towards a world where accents matter less than fixing accents. Sure very thick accents can hinder communication but I feel like improving that a bit is enough. There's no need to perfectly emulate the native accent (and let's not forget about the differences between say Jersey, Oklahoma, Boston and L.A.).

Secondly, I'm not sure an app is really needed for this. Someone dedicated enough to install an app for this might also be dedicated enough to turn on Netflix and just speak along until they sound similar to the actors.

Either way, good luck to the team.

Hey, thanks for sharing your thoughts.

When it comes to striving for a world where accents don't matter, we feel the same way.

In fact, our dialect coach, Ron Carlos, slacked me earlier today: "We truly hope that one day accents won’t matter, but until then we have folks who feel embarrassed about their accent which keeps them from showing up with their full selves. We’re here to help those folks feel more confident with their speech."

What we're trying to help users with is learning the physical skills that make up their account: pronunciation, speech rhythm, intonation, stress -- ultimately, how to speak the way they want, with the ultimate goal of helping the user become more confident and clear in their speech. If the way they want to sound is exactly like someone from Jersey, Boston, L.A. or anywhere else, we're happy to support them!

> First of all, it feels like a patch for latent discrimination.

Try learning a language that's not English. You're definitely judged on your accent, ie. your ability to properly pronounce words. If you want to fully assimilate to the environment you choose to exist in, speaking with less of an accent helps with being understood. Also, the 'standard' American accent (ie. Hollywood/West Coast accent) is understood by absolutely anyone; even those who are new to speaking English.

Heck, even English speakers from regions such as the Caribbean, England, Australia, etc..., turn their accent 'off' in business environments to be better understood. Many native English speakers can't even understand creole from the Caribbean for example.

Accents can definitely also hamper communication. I'm a native English speaker who's interacted with many immigrants so I can definitely understand almost every accent. But two immigrants both speaking English as their second language and from different parts of the world will often have trouble, I've seen it plenty. Even my girlfriend, who speaks English at quite a high level but isn't native, has a difficult time if I speak too quickly using local slang and pronunciation (and our accent barely deviates from standard American English).

Focusing on a standard pronunciation, especially in a job or school environment, helps a lot.

I wouldn't really call it discrimination if the dude on the phone has a hard time deciphering what type of sandwich you are trying to order, or talking to a stranger in a loud bar who needs you to keep repeating yourself.

I think the need for this stems from a desire to be understood. Though I understand why the OP would list discrimination in their sales pitch.

I'm a native English speaker, but I'd love something like this for French, Spanish, or Mandarin! Any good products you're aware of?

One of my coworker and good friend literally named NPCs in the game we were developing together at work after my pronunciation mistakes: kunter (counter strike), eevol, etc. In French we have an 'ee' sound you dont even have in english, it's much more 'ee' than ee. Took me years to hear the difference. Pretty sure it is one of the hardest spot for a french: beach, snitch, sheep, cheat, etc. And none of those sound like an i in french (e.g frite).

Snitch doesn't sound like beach, sheep, cheat. I'm not sure how to approximate snitch in French, maybe you can't. The Google card that comes up if you search "snitch pronunciation" is right though.

Amazing product!

Congrats on the launch! Worked on something similar, so rooting for ya. Using Hollywood accent coaches is a great approach, but I'd also look into identifying speech therapists that specialize with helping adults in target regions to 1) get more information for your models and 2) point out which problems can't be corrected in older adults, so you focus on those that can.

When you succeed I hope you expand to other English accents, e.g. Australia, NZ, England…

This looks interesting and as a non-native English speaker, I'd be open to pay for it if it fulfills it promises.

However, the fact that it is a mobile app only is quite annoying. Why can't I just use this on my laptop? Some of us are actually phone-less :)

I _really_ wish this existed for other languages and other English accents. I'd love to use it for that!

I'm Bruno, founder of Abstra (https://abstra.app/). Abstra is a no-code tool that allows designers to build web apps. Imagine Figma, but we export a functional web application. Here's a timelapsed demo showing it in action: https://www.youtube.com/watch?v=5fObVcHcd_w.

As opposed to most famous no-code tools, we don't do any nested-based editing (e.g. tree structure like Webflow - which is an amazing tool, but is basically a visual HTML editor requiring almost the same skills). Instead, we have free canvas-based editing, so we let designers freely iterate as they do with prototyping tools like Figma/AdobeXD/Sketch/etc.

Our visual interface builder lets you add controls using "springs". This is a kind of constraint layout (like XCode/Android studio does) without Cassowary solver (which means more tolerance to inconsistencies). This idea actually came from a master thesis project with my girlfriend where we created methods to simulate elastic material deformations. Then I realized that those abstractions (elastic springs) are easier to understand than the flexbox/grid/box model for UI layout.

Abstra is great for companies that want to make MVPs in production, learn with real data and move faster. We already have a few companies using Abstra. One of them is Stone (Nasdaq: STNE), which uses us to build internal tools faster than coding by hand. Examples of what people have built so far include a customer support portal, a field sales collaborative notes app, financial controls, timesheets, an employee evaluation tool, and a student enrollment control.

I’ve been dying for an AutoLayout for web — can’t wait to try it.

Just curious - at what point does someone “grow out” of this tool? When you say Abstra is great for MVPs in production, I assume there’s a next level not served by this tool?

Great, thank you!

The next level is "complex apps" because code is still a better way to create customized abstractions.

That said, you can scale simple/CRUD applications with Abstra for tons of users without having to mind about infrastructure

Isn't AutoLayout in Figma just a partial implementation of flexbox, an therefore already exist on the web in a more complete form?

This looks pretty neat. Just wanted to give you a heads up that the word "regulamentation" might be a typo in your docs here:

https://docs.abstra.app/#security-by-default

Even outside of that word the paragraph itself reads kind of weird:

"The Abstra team makes all the necessary effort to make your project secure and compliant with all the actual rules, good practices, and regulamentation."

I took a stab at changing it:

The Abstra team is dedicated to securing your project in compliance with rules, regulations and standards using industry best practices.

Not sure if that jibes with the legal side of things but I think it reads better.

That does read better, however I like the portmanteau of "regulatory + documentation"== 'Regulementation'

The NSA hardening docs could count as such...

Thank you! We will change that!

Whats your email plz (put it in your profile)

Good call! done! bruno [at] abstra.app

> Then I realized that those abstractions (elastic springs) are easier to understand than the flexbox/grid/box model for UI layout.

I think so too! I enjoyed using Bootstrap precisely because my mind treated those auto-layout 12 columns as equally elastic springs, but this abstraction leaked very frequently.

Really cool to see such notion embedded into a layout builder.

Thank you! I think there always an abstraction bias that makes webdev frameworks so coupled with HTML.

Great to hear that you liked!

This looks really cool.

Thank you!!

This looks amazing! Curious to know what stack you are using to build the frontend. I am starting to build a video editor, so just curious if any framework (React, Vue) or library were especially useful?

Thank you! We are using Vue+Typescript for the frontend and Haskell & Node for the backend microservices.

For your case (and maybe for us in the future), I would take a look at rust+webasm or using WebGL directly

Thanks, currently I am also using Vue and doing simple DOM manipulations.

Any specific Vue libraries you found useful for drag-n-drop, alignments, connecting arrows, etc?

We saw limitations on all libs we tried, so we decided to implement drag-n-drop by ourselves using setIntervals, getBoundingRects and absolute positions for arrows and other visual elements

amazing product! i'm looking forward to test it

Great, thank you! If you have any questions or thoughts about our tool, you can contact me (bruno [at] abstra.app). I would love to get your feedback on it.

Congrats on the launch. Have you tested the generated UI for accessibility, particularly with a screen reader? If I can try a finished demo app without going through the design process myself, I'd be happy to provide feedback.

Yes! We have a customer which have blind users. So we are prepared for this!

Here is a quick demo: https://accessibility-demo.abstra.app/

Not bad! The only slight problem I noticed is that NVDA (a popular open-source screen reader for Windows) announced that the plain text above the edit box was clickable.

Excelent feedback! We will fix that!

I couldn't find any live demos of webpages built with it, is there any examples?

Hi, we are Salvador and Max from Levo (https://www.levo.mx) a high yield savings account for Mexico. We aggregate the savings accounts of our users and negotiate significantly better interest rates from other banks on their behalf.

65% of Mexican savers keep their savings in their checking or savings accounts, yet 97% of all bank accounts are non-interest bearing. This is partly due to a highly concentrated banking market, where 7 banks hold 80% of the market.

I (Salvador) walked past the CD-Pricing Desk at Mexico’s biggest bank every morning for 5 years. In 2017 Mexico finished the year with the highest inflation rate in two decades and the Central Bank’s overnight rate was 8.25%. But Mexico’s biggest bank offered clients rates below 3.5%. This is when I realized how broken the system was and set myself to find a solution to this.

Levo is a certificate of deposit marketplace. People who save money in their checking or savings account, can access CDs from multiple banks through our app and get up to 300% better rates. For example, our client Gustavo recently got a 50,000 USD inheritance; his bank offered him 4.5% p.a. for his savings, with Levo he got 6.5% p.a. with one of Levo's banking partners.

We are currently running a private beta, if you are a Mexican resident please reach out to be part of our special group! If anyone has questions or comments we'd love to hear them!

Very interesting. Does Mexico have anything akin to US Federal Deposit Insurance? Does Levo?

Yes! Its called IPAB and its part of the federal government.

This idea looks good. Finance revolves on information asymmetry and keeping your customer uninformed on what's best for them.

What does the final step on account opening and transfer of funds look like for the user? Surely banks can't be that cooperative in providing you an api to open accounts for them.

Are there any money market funds or short term debt funds in Mexico?

I am sure you would look into adding forex cards and remittances in future. There are a couple of startups here in India which help you get better forex rates by negotiating with banks for pre loaded foreign currency cards, hard currency and TT/Swift transfers.

In terms of using the App, once we get approval from the regulators, users just have to download the app and do our onboarding process, then wire their respective funds via SPEI. And yes indeed, most Banks don't have APIs, however our onboarding process is compliant with official KYC requirements and that's how we will be able to operate.

You mentioned an example of a $50k USD inheritance getting 6.5% APR, is that a peso denominated account, or a dollar denominated account?

As an expat living in Mexico as a permanent resident, a dollar denominated Mexican bank account that pays above inflation rates is very appealing.

The account is peso denominated. If your are interested we are happy to make you part of the private Beta.

Growing up in Mexico and now working in the financial service industry in the US, I’ve seen how many of Mexico’s residents are under banked or unbanked. I think this is an important issue you are trying to solve for! Is there a direct link I can share with family/friends in Mexico to participate in the private beta?

How does this compare to the end user with holding USD? I guess you'd get hardly any interest in USD, but at the same time much less inflation.

I already get 6.5% from HeyBanco. Actually, they already increased it to 7.25%, and keep in mind these CDs expire every 7 days so the money is quite liquid. How are you any different?

Hi HN, we're Karina and Andres of Monto (https://www.monto.mx/). We give employees of our affiliated companies access to their earned wages 24/7. We're currently operating in Mexico.

With Monto users can pay for basic services (like water, electricity) on time, avoiding penalties. They can stop paying overdraft fees and avoid having to resort to abusive credit alternatives like shark loans, employer loans, payday loans, credit cards, bank loans, department store credit cards, pawn shops. The conditions on these options are abusive and prohibitively expensive for low-income workers and have resulted in serious financial stress for millions of people.

Andres was working in investment banking in New York, but ended up back in Mexico for a while working on the family business and advising some friends with their business. Through these experience he was shocked to find out: 1) The level of indebtedness that low-income workers had in Mexico. 2) Companies were not offering benefits that promoted financial wellness in most cases they were offering problems that exacerbated the problem. 3) The lack of transparent, low cost, easy-to-use financial products in the market. He decided to move back to Mexico after living abroad for 12 years, and start a company that would fix this for people.

Users have immediate access to their earned money with Monto - we deposit the money in 2 minutes into any bank account. Withdrawals are secure and confidential. Users don't need to fill out an application, provide documentation or fulfill a long list of requirements – any employee of our affiliated companies can use it. Users pay a small fixed charge per withdrawal, regardless of the amount. There are no interest payments or hidden fees. In contrast to other alternatives, we don't perform credit checks and usage of Monto cannot affect our users in any way (e.g. credit score).

> Users pay a small fixed charge per withdrawal, regardless of the amount.

Really disappointed to read this. Wages are so low that working people can't pay their bills, so your solution is to lower their wages even more? There was another YC company like this but they were charging the company, not the employees. You may not be a loan shark, but you are enabling the company to pay low wages and offloading the cost of mitigating the problem to the already-exploited. This is ethically dubious, at best

Agree with you. If Monto only works with 'affiliated companies,' at least strike a deal with those companies to take the financial burden off of their low-wage employees. That other YC startup Payflow (that was actually ripping off UIs, to their detriment) proved it to be possible.

Employers or users can pay the withdrawal fee, that's up to the company. We encourage companies to offer it as financial benefit. Regardless, users love our product and see a lot of value in getting the flexibility to decide when to get paid. In contrast to other options in the market there are no interest payments or hidden costs.

I can see how in the short term it's a positive service - they can now keep their lights on, their water running, etc.

But as a long-term proposition, a permanent adjustment to the status quo, it just feeds into an exploitative system in the same way that loan sharks do. You're filling a gap that should be filled with higher wages, and reducing wages from the most-vulnerable to do so.

I agree with your sentiments, loan-sharks gunna loan-shark....

Though, if there is a good lining here, its not the retarded interest rates check-cashing and pay-day loan places charge... and thusthis seems a step in the right direction...

Although, I am also reminded of MONDO.COM

YOU CAN DO ANYTHING AT MONDO.COM!

THats what this BSP sounded like...

----

I wish them well though.

You're arguing for reducing options for people who don't have money to ... make yourself feel better? Surely them having an option to choose for themselves is better than no option, unless you think you know better than them (which would be pretty arrogant and small-minded of you IMO).

To be fair to the GP, your video does say "Sin costo para su empresa", so that sounds like the route Monto is pushing.

I think it is very predatory. You guys are "changing the world" (for the worse). We should encourage people to save not to borrow money with interest and spent. They will get poorer and poorer and can never get out. You are taking advantages of poor people. So I hope you fail.

People who are not making enough money to make ends meet -- and thus end up with overdraft fees or having to make use of things like instant loans or credit cards -- don't need to get paid faster. They need to get paid more.

Whether they can afford to live or not is a matter of simple arithmetic: money left = money made - expenses. Being able to tap into their underwhelming income faster to cover a mid-month bill doesn't help them in any way, it just pushes the problem around, so that come payday, they will have even less and have to repeat the same process.

This product is only less distasteful than text message money lenders because it doesn't come with exorbitant interest rates. This product isn't helping anyone. It's just capitalizing on the poor.

> People who are not making enough money to make ends meet [...] don't need to get paid faster. They need to get paid more.

Or they need to pay less. If the cost of being poor represents a significant fraction of the total spending for these people, lowering that through arbitrage is still a benefit.

I mistakenly went negative once on one of my bank accounts (purely due to sheer laziness/inattention), I was amazed by how expensive that small mistake was. People who are unfortunate enough to have than happen regularly spend a lot on those fees, especially compared to their almost non-existent disposable income. Anything that can help/reduce those expenses is a win in my book.

You're neglecting the fact that they will also have expenses the month after, and the month after that. Tapping into their income faster just means that this month's bill gets covered by their salary, but next month's bill goes on credit.

If anything, this product is bad like credit cards and overdrafts, in that it will enable people to make financially unsound decisions while feeling like they can actually afford it, when they can't. They're just burning the candle at the other end.

For ease of numbers, let's say you make $100/week and have to spend $90/week on personal expenses.

If you get paid once every 2 weeks, you have: Week 1: -$90 Week 2: +$20 Week 3: -$70 Week 4: +$40 Week 5: -$50 Week 6: +$60 Week 7: -$30 Week 8: +$80 Week 9: -$10 Week 10: +$100

So it takes you 9 weeks before you never go negative and to be able to afford to take the job, you need $90.

If you get paid everyday, it's: Week 1: +$10 Week 2: +$20 Week 3: +$30 Week 4: +$40 Week 5: +$50 Week 6: +$60 Week 7: +$70 Week 8: +$80 Week 9: +$90 Week 10: +$100

It's clearly nicer to get paid everyday all else being equal.

Yes, you're absolutely right. If we set up a perfect example to support your arguments, and we ignore any possible negative aspects, then we can conclude that this is a purely beneficial arrangement.

Look, I'm not arguing that it can't be useful to have the ability to get paid daily sometimes, but I don't buy into this schtick about this company doing this in some attempt to help the poor, because literally the opposite is happening.

A more realistic picture goes something like this:

In the middle of the month with only $200 in your account, you get hit with a $400 car repair bill. Instead of having overdrafts or text message money lenders, you can now borrow from your future salary via this company. You do this, cover your bill.

Now it's $next_month and you just got paid $400 less. If you were $200 in plus in the middle of last month, it doesn't seem far-fetched that you'll be in a similar situation this month, which means you're already at -$200, but you just don't know it until the middle of the month.

So what do you do now? Do you borrow from your salary again and repeat this process? Maybe, but then go back to step 1.

An interesting related tidbit is the fact that poor people are actually more likely to make bad financial decisions(and health and etc) and there are numerous studies showing this.

The only thing this product is doing is handing the poor yet another instrument to create financial trouble for themselves, under the guise of helping them. I can also frame it like this: How is this any different than if this company instead just offered "no nonsense loans" at only a 2% interest? Hint: There is zero difference, except in terms of risk for the company. The genius of this arrangement is that they never have to worry about people not paying off their loans.

I think your last point makes your true beliefs clear. You believe it's better for the poor to have a forced savings account (or as I call it a loan to the employer) so that they can't "make bad financial decisions". It's extremely paternalistic to believe the poor shouldn't have immediate access to all of the money they've earned or at least a choice of when to do so.

Of course there should be legal limits as to what interest rate can be charged. I don't believe anything more than 10% on an annualized basis can be justified.

In any case, the vast majority of your points in this discussion are irrelevant and the most potent point you make seems to depend on either interest rates not be fair or the poor being too irresponsible to get their deserved cash.

No, I think the poor should have as many choices as everyone else.

My "true beliefs" are that this company isn't some kind of altruistic attempt at "helping the poor" - this is just a way for this company to capitalize on the poor in the same way any other predatory payday loan company might, except they can do it without running the risk of not getting paid.

I think this is a pretty extreme take and probably not fully warranted. I do agree that the bigger problem is enough money not the same money sooner.

It is kind of messed up though that you put in work and then your employer decides to pay you 1 or 2 weeks later. As an employee, you're essentially giving a short term loan to your employer. You also carry the risk for whatever reason they wouldn't be able to cover payroll. Floating a week or two of personal expenses may not sound like a lot but if you're living pay check to pay check it can be a ton.

Is this idea going to significantly change workers' rights and standing? Absolutely not. Is it poverty profiteering? Probably not.

Hi - we’re Alex, Thiago, and Marcel, founders of Plug (https://www.plugpagamentos.com/). We are live in Brazil, and currently onboarding our first beta customers in the US and LatAm.

Plug is an API to manage multiple payments providers. Through a single integration, businesses can connect to multiple providers and route transactions between them.

It’s hard to know who the right payments provider for your business is. In the early stage you may care most about a fast integration, later you're looking for lower fees and more payment methods, and finally for better acceptance rates and fail-over options. We built Plug so companies could focus on their core business. With our multiple connections, provider-agnostic vault, and routing engine, our users always work with the best provider, on a transaction-by-transaction basis. Working with multiple providers means you can optimize costs and acceptance rates through routing and retries, and that you will be able to avoid lock-ins.

Brazilian merchants in particular change their payment provider frequently. They do this to support multiple local payment methods, to try and improve the worst acceptance rate in Latin America (77% according to Visa), and to reduce their costs, which in some cases can be 10% or more of the transaction.

Alex was at Braintree, Marcel was at Gympass, and Thiago was at Medicinae. We got the idea from seeing our best clients, companies like Uber, AirBnb, and iFood, build their own payment abstraction layers to work with different providers. We've built software that allows any business to do the same: one integration for multiple connections, and the ability to always transact with the best provider without having to write extra code.

Rei do Pitaco has been using Plug to reduce their costs and support multiple payment methods through different providers. CartX has been using Plug to offer multiple providers to their client base without having to integrate each provider. Simplix uses our product to optimize costs by routing to the least expensive provider for each transaction. We're looking forward to hearing your comments!

I've got a couple of questions: Do you plan on developing a no-code interface soon? Before learning to code, Stripe on my own landing page was a no-go. Do you help in completing the integration after your own or should I do a proper Stripe integration after Plug?

Congrats on the launch anyway, I'm convinced that your solution makes sense for early stage company. You'll have to provide more value in the future to keep enterprises in the long run I guess.

Hi Arthur - thanks for the questions. Our checkout option is embeddable, but does require coding. We will be releasing payment links, which is essentially a landing page. Might be a good option for your use case. As for the second question, there is no need to complete the Stripe integration after integrating Plug. Once you've integrated our API, you are able to use any payment provider we have integrated. Enterprise companies see a lot of value in the platform actually! It allows them to have their own, provider-agnostic vault, and the increase in acceptance rates through retries and routing significantly impacts their topline. But completely agree that we consistently need to provide more value for all clientes...independent of size :)

feel free to reach out if you have any more questions: a@plugpagamentos.com

Could you give us a high level breakdown of your KYC/AML compliance?

Sure! We're actually a software layer. The KYC/AML is done by the providers. We only help our clients ensure this is a smooth process by letting them know the steps beforehand and aggregating all the relevant information before sending through.

Even in the USA Bodegas have a huge problem with card transaction fees, for example, the one near me charges $.80 per card transaction... So, imagine if you are buying something for $1.60 on your card (Many people do this, especially when you observe them using their unemployment/SNAP (Food Stamps/Government assistance cards)) -- so they pay %50 of the transaction just to swipe their card....

This space is very red ocean. What makes your product unique than other more mature offerings?

I don’t see how this solution is in a red ocean? Mature offerings where? The only non-Latam alternative that comes to my mind in this space is Checkout.com and that is just partially similar.

A provider routing middleware with the goal of increasing authorization rates is likely not even an attractive business in markets like the US and Europe where you probably can defer to a single payment provider like Stripe for maximum checkout reliability.

This is not the case in Latam. Authorization rates are way lower and anti-fraud heuristics and models by local payment providers are less sophisticated. Most growing business in Latam end up in a situation where they neeed to re-route payments to increase their authorization rates and sometimes in the need to build very strict checkouts that harm conversions but that are needed to avoid a high volume of rejected transactions.

Definitely a market need for this, having built one of those layers myself.

Are you planning to support saved payment methods across multiple providers? Thats always seemed like the touchiest area for me, especially since it would seemingly invalidate a lot of the fraud protection that payment providers use to keep lock-in.

That's awesome that you built that layer yourself. If you're up for it, would love to connect and hear why you did so.

We're PCI compliant so can use saved payment methods across multiple providers, however in our case it does not invalidate the built-in fraud protection that providers use. Additionally, you could also use third party fraud providers at any point within your processing flow.

It was for a big subscription service and we needed to enable a lot of promotional capabilities and also keep payment provider costs low + be able to operationally switch over in high traffic times.

> however in our case it does not invalidate the built-in fraud protection that providers use

Thats pretty neat! Even Stripe Radar? My understanding was that used their client side fingerprinting when the card is inputted, but its been a little bit.

Reply on news.ycombinator.com