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!

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?

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.

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.

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/

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.

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"?

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

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.

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.

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!

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 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.

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.

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

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?

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.

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 :)

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.

We are a diverse bunch

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

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

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

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

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

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.

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.

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

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.

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.

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.

>, 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.

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.

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.

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

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

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.

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.

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.

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.

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.

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.

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?

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

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.

Great to hear that you liked!

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

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

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.

The NSA hardening docs could count as such...

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

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!

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.

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

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).

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

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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!

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.

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.

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.

> 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.