The privacy and security part is not inspiring confidence. Scrolling to the next section got me thinking "Don't get scammed at closing, get scammed before closing after uploading your mortgage documents to a random website."
Cool idea though.
Hey, Aaron the builder here.
The scamming that happens to homebuyers is not even comparable to the risk in uploading docs to a website which promises they won't share user data with anyone. This is genuinely a pro buyer tool with no association with any 3rd party.
The tool has already helped many people negotiate and get a better deal on their mortgage. Please before judging understand that 70% of buyers overpay in their mortgage 1-3% in closing costs and bad rates. It's mind boggling how much lenders get away with profiting in junk fees from stressed out homebuyers.
Allow me to expound on @kojeovo's remark. Please take this as a constructive criticism to improve your success potential. Much of it is from a quick glance, and am sure there are many other facets to improve.
A business is not just about the product.
Your Privacy Policy. There is no default way to download it (see 9.), and since it is window-ed cannot print entire doc. That means I cannot keep a copy of it for myself.
> We collect the following types of information:
> Mortgage Documents: Loan Estimates and Closing Disclosures you upload for analysis.
Okay, but
> 4. Data Security
> We implement industry-standard security measures to protect your information from unauthorized access, disclosure, alteration, and destruction.
This means nothing. Are you ISO 27001:2022, NIST SP 800-53, CIS, CE+, Essential Eight, or something else? Have you been audited, and proof? Who is your ISP? What regs do you follow around data sovereignty?
Terms of Service. Again, no default way of download. Overall, I would never agree to this ToS. It demands all kinds of requirements on the user, but takes no responsibility for anything - or as described above, explain how you will protect your customers.
You have no reference anywhere where you are geographically. No address, no about us, no who you are. I would be very leery on uploading anything.
Thanks for the constructive feedback.
I just added a way to easily download the entire privacy policy and terms of service, also quickly added an about page with some info about me - https://closing.wtf/about
Eventually I'm going to get a certification and will keep your other points in mind.
Would it matter if they had a "perfect" privacy policy? I don't believe there's anything legally that enforces it. So they can promise the moon then turn around and sell your data.
Maybe I'm wrong here, but, My mental model of privacy policies and the like has always been: This is a lie, the company will do whatever it wants with my data. And I will have no recourse.
As such I've always acted accordingly. And very few websites have legit info on me.
I think acting 'as if' is the safe option here but encouraging change for the better in someone willing to engage in dialog is still better than not doing it. Maybe you didn't intend to make a counterpoint, i just wanted to point that out.
Legitimately curious, what’s the worst they could do with this data?
The most common scams around home buying are wire fraud - contact the buyer pretending to be the title company and steal their money. The data in a mortgage is exactly what you need to enable these scams and you're getting people to hand it to you and at the same time tell you they are about to wire money.
Wire Transfers are not undoable and instant, much like Zelle. So I always recommend people send $10 first, and confirm everything works, before sending real money. When doing the confirmation, try using a different channel of communication, to ensure you are getting the right person. i.e. call them directly from known good phone numbers or something.
Yes many banks charge $30 or more for a wire transfer, but I'd rather just pay the $60 than have a large sum wire transfer lost, stolen, etc.
Some banks/Brokerages are sane and do not charge extra for wire transfers. Fidelity is one such. BOA also(if you have enough assets there, $100k will do it).
Yep. When we closed on our house we got a whole lecture from the title company about how frequently data breaches lead to wire fraud and to not trust anyone. Mortgage originators are constantly under attack to try to get at the information that OP is asking people to just casually upload.
Their aggressive dismissal of the concern is not a good look.
I am not dismissing the concern, I was stating the tool solves an even larger concern. I'm doing everything I can to setup it up to be secure, private, and worthy of trust and addressing the feedback points.
If you have suggestions more than "don't trust this random internet tool even if it gives you free advice, regardless of the value it offers", please let me know [thanks emoji]
I have never done a wire transfer at a residential closing. I come to the closing at the title company office with a cashier's check from my bank for the amount they told me to bring.
Cashier's check was only accepted for amounts less than $10k at our closing by our title company. This seems common to require wire. The title company contracted with a 3rd party escrow service so the money was required into the title company's account at the escrow service. I assume a cashier's check would need to be mailed to the escrow company
Did you bought enough houses to assume that's always the case?
My experience was that I was told to send the cashier's check using overnight FedEx because they did not have office in my area.
No, fair enough. I would not close anywhere other than a local title company though. I've had a few odd things surface at the last minute that were resolvable because everyone was sitting around the same table.
The only method available to me at closing was a wire transfer. It is dumb.
Aside from the personal details (name, address, etc), they can collect pricing info on houses, run analytics, and swoop the deal with a slightly better offer or better yet, sell it to wholesale buyers, reits, and whoever is interested in stealing the deal.
>they can collect pricing info on houses, run analytics
AFAIK house sale prices (ie. property transactions) are open in many (most?) jurisdictions.
>and swoop the deal with a slightly better offer
How does that even work? The winning bidder is presumably someone who gave the highest offer. Why would another company pay above and beyond that, considering that there's probably several other serious buyers who aren't willing to pay more?
The terms are not public until the house is sold. In the contract pending state you don’t know how much it is going to sell for. Theoretically if they saw a buyer accepting a crazy low offer they could alert the troops.
But it doesn’t need a lot of the data in that document, so really they need a way to redact all the unnecessary data to require less trust.
Edit: words.
The deal isn’t always about the price. For example, a $1M house bought with $100K down and $900K mortgage is a worse deal for the seller as compared to $500K down and $500K financed. Assumption here is that it is more likely to get a $500K loan irrespective of the appraised value of the house.
A lower all cash offer (say $975K) is likely a better offer for the seller because it reduces the risk for them and closes the transaction much quicker than a mortgage transaction.
I have been a buyer in two transactions where my offer was slightly lower than the highest bidder, but with better terms.
> a $1M house bought with $100K down and $900K mortgage is a worse deal for the seller as compared to $500K down and $500K financed.
Do sellers in the US know how large your down payment is? AFAIK that's not a thing in Canada. Offers either have a financing condition, or don't. If the offer doesn't have a financing condition, the buyer might be paying cash. But they could just be trying to present an offer with better terms, gambling that they'll definitely find financing somewhere or the other.
Yup, the seller is (at least should be) made aware of the financing structure, as it’s part of the offer.
Every time I’ve sold a house it’s been a factor in deciding which offer(s) to pick or counter.
As a home buyer, I've been beaten many times by an all-cash offer that was significantly lower than my financed offer. For example, a $450K all-cash offer where they'd close in 7 days beat my $525K 80/20 offer where it would have taken me 25+ days to close.
> name, address, etc
In my country all that plus your social security number and tax declarations etc are public information. What's your opinion on that?
That...is not how mortgage servicing companies operate.
People aren't concerned about giving their details to a mortgage servicing company, they're concerned about giving their details to a random website called "closing.wtf", which promises to provide mortgage advice for free with no other obvious revenue source.
Lol no revenue source and a promise to never sell or share their data.
Gotta figure this one out...
FYI, this reads as a very aggressive response to someone raising legitimate privacy concerns and doesn't engender the trust you very likely deserve.
Rather than talking up the value of the tool as superceding the concerns, a more constructive approach might acknowledge the concerns and emphasize how you already do minimize risk or commitments you're willing to make towards doing so.
Being dismissive doesn't help worried or skeptical people feel more secure, and worried and skeptical people make perfectly good users too.
It is fair to describe the pains of not getting analysis on mortgage loan estimates, but what I think folks are looking for is some kind of authentic answer to the problem posed.
For example, you could advise the person uploading to remove PII prior to the upload, and link to pdf editing tools that allow them to do that.
You could say that not including PII like full name(s) found on just about every loan estimate does not take away from the value of the tool.
Another thing that could be done is to provide clear means for removing any data uploaded, or opt-out pre-upload of any data being used for training.
For example by creating an account first.
Providing some skin in the game such as putting the removal behavior in the terms of service and a personal guarantee to do everything to ensure sensitivity to privacy of this information will be handled carefully staking your reputation, probably would help.
I am genuinely surprised by the comments in this thread.
Privacy concerns are real but the importance of that matter in your project is overestimated here by an absurd level.
What I read is not a constructive criticism and the suggestions laid down are not realistic nor business relevant at all. I feel like this is some sort of mass wishful thinking.
I think it's actually refreshing to see the top comments and constructive criticism be about privacy concerns. It shows that even for little "Show HN" projects, there is growing intolerance of half-assing it. Not saying OP in particular is half-assing it, but it's good to see these questions being regularly asked front and center. I honestly wish the Tech Media paid more attention to privacy and security instead of just copy-pasting companies' PR statements as "articles."
Title deposit wire fraud is a very big risk. The amounts are devastating to the victims, so the operator has to go above and beyond to secure the data because of the huge risks involved. Would you risk losing a 5-/6-digit amount to fraud in order to potentially save on a 4-digit closing fee?
> Privacy concerns are real
This isn't about privacy, it's a security concern. People's life savings are on the line here, and the information OP is requesting is enough to pull off very sophisticated social engineering attacks. It's entirely reasonable to ask what they're going to do with that information and how they're keeping it secured, and their reaction to the questions is entirely inappropriate for someone who's asking for this degree of trust.
> not even comparable to the risk in uploading docs to a website which promises they won't share user data with anyone. This is genuinely a pro buyer tool with no association with any 3rd party.
I have no reason to think you're not completely sincere in this!
But, realize it doesn't mean anything.
Unless that promise is backed by some ironclad contract, it means nothing. Companies grow and hire new people who don't care about the original values. Or they get acquired and all bets are off. Or they start running low on cash and suddenly decide monetizing all that data is a good idea after all. Or it becomes visible enough to attract attention of the government who shows up demanding copies of data. And so on.
I've been in one or more startups where all of these things have happened.
How can you get scammed on a mortgage? They're typically standard products from nationwide banks.
Great idea and execution. I understand the privacy concerns, but I believe implementing a client-side redaction step could alleviate some of them. This step would allow users to preview their uploaded content before submitting it. While designing this feature, it’s crucial to ensure user trust and convince them of its benefits. Personally, I would feel more comfortable uploading a PDF knowing that it will be anonymized or redacted before being submitted.
Doesn't matter your promise, even though you may or may not be trusted, hackers can get it and steal it all. So it's not necessarily you or your service.
Ignore the haters, they will probably never be your customer.
People are trying to increase the potential customer base of the author by pointing out where there is room to improve. That is incredibly valuable, and one of the major reasons to do a Show HN.
That is not being a "hater".
The percentage of regular people who care about any of the risks discussed in this thread is approximately zero. For better or worse.
Your typical home buyer isn't reading the contract they sign when they buy a home, let alone the privacy policy of a simple tool they use to check if they have a mortgage with decent terms.
When people who do care about privacy and security make their voices heard, such as the case here where the owner has committed to improving their policy & processes, it benefits everyone using the product or service.
It’s not in the spirit of a “Show HN” to attack the guy’s privacy policy and shut down the idea over security concerns that the general public wouldn’t think twice about.
No one "attacked" the author.
The first comment said "not inspiring confidence" and then WaitWaitWha gave a very thoughtful comment with actionable advice, based on that comment and the reply. These are not attacks.
https://news.ycombinator.com/item?id=42150219 was highly constructive. It was direct and actionable.
kojeovo's original comment was less so. When you build a product, you're going to get random, in-actionable comments from people who just like to complain. Separating the signal from noise is difficult, and while there is a underlying concern about privacy, not giving anything actionable moves it towards to the noise side of the spectrum.
Absolutely agree.
Neither are "haters", though. And, speaking on quality of feedback, "ignore the haters" seems fairly low.
Yea, "ignore the haters" is terrible advice. It basically means "stay in a bubble where the only feedback you listen to is positive".
I won't even get into how ridiculous it is to consider anyone who disagrees with you a "hater."
Ah the Disney approach.
Bold strategy Cotton.
Owner did the smart thing and listened to the constructive criticism which made me feel infinitely better about using his tool.
Which I will now do, and would not have before. I am also his exact customer.
Also, you have no control over decisions that any future owner might have, and you won't care because you've already cashed out.
What happens when you get hacked? Not if. To come back at someone with valid concerns with a "no, you don't understand my point of view" does nothing but a disservice to you.
Expecting people to just accept things is just not a good way to operate. When you receive push back, you need better responses than this. Will the vast majority of your users push back, sadly, probably not. However, you did post this to HN and then reacted poorly to valid criticism. Tsk tsk
I love this idea (haven't tried it) and it seems like a killer app for AI. I can think of a lot of other things like health insurance, home owners insurance, and many other types of contracts for which an AI advisor can be built for. Imagine being able to rake over a complex document and make decisions that clearly benefit you. That's a rare privilege.
>The scamming that happens to homebuyers is not even comparable to the risk in uploading docs to a website which promises they won't share user data with anyone.
Well as long as you promise, my privacy fears are allayed!
/s
What sensitive data do you think is on a loan estimate? I've received a dozen over the years and it's literally just your name and the address of the property you want to buy. Both of which are public information if you do purchase the property.
Public info that you bought a property is entirely different from info about all the properties you are searching for and seriously considering. Especially being able to couple that with what you eventually did later.
If you're uploading a loan estimate you've already made a loan application for a specific house- this isn't going to give them information on every house you were considering, it is probably just the single house which accepted your offer and are shopping around for financing.
Why would a name and a random property address have value before the transaction, but not after it?
Well, one is before they spend a lot of money, and the other is after. In only one of those scenarios are they prone to spear-phishing wire fraud on the Title deposit. The victim is already primed, and has indicated they have gobs of money set aside.
As others have mentioned, title fraud. My recent closing disclosure has buyer, seller, agent, and title company. It'd be pretty easy to call the buyer claiming to be the title company and request a wire for exactly the right amount to a fraudulent location.
We're talking about enough information to create a list of people who are in the process of making the largest purchase of their lives, coupled with the name of the businesses that they're considering sending money to and the exact dollar amounts that they're considering sending.
Scammers pay a lot of money to people who can get them those kinds of lists.
I once knew a founder of a pre-GPT-3 AI product that analyzed certain cost-adjacent documents to find "hidden" optimizations. The "AI" was the founder, an expert in that industry, churning through the uploaded documents himself and writing reports by hand detailing potential cost savings. How far we've come!
Some people suspect that is still how many of the AI startups operate
If there are only a handful of operators in a given industry and they use the same billing format, why not?
Create a known good OCR to calculation mechanism, then generate reports based off it. If it is inaccurate, its probably a small amount of logic to fix it.
With GPT you could even get it to write the parsing logic for you perhaps, and maybe process bill data when a bill doesn't exactly match existing parser data.
Reminds me of Mechanical Turk's old tagline: "Artificial Artificial Intelligence"
This is great. Really good work. This looks really useful, and I think you did a great job on the interface and site in general. I hope you aren't discouraged by the other commenters, keep hacking!
Was going to say the same. I ran into the same problems when I bought my first house and spent tens of hours trying to untangle and reduce costs. Thank you for tackling this
Your welcome, thank you for the message.
Thank you!
I tried it out with docs from my last refi. Tried both PDF and PNG files, both returned "An error occurred with processing your document. Please confirm you uploaded a valid Loan Estimate or Closing Disclosure PDF."
I also at first didn't notice there is action needed to choose between uploading a Loan Estimate vs Closing Disclosure. It doesn't seem insurmountable to have the site automatically figure out the difference between those two.
You're right. I just haven't gotten around to figuring out the document type automatically even though it should be straightforward to implement.
There are many reasons the document will fail processing, it's usually because the wrong document type was selected or the original pdf from the lender has been modified. Sometimes, users upload a document they received which is just random text in a pdf they received as a pre-offer from a mortgage broker or lender.
This looks really useful. I'll definitely be using this in the near future. However, I unnecessarily downloaded the sample loan estimate PDF and uploaded it again in order to view the report. Please make the "view sample report" button more prominent, or add another "view sample report" button within the blue "Try it out!" box.
oh a cool and useful idea on HN, i'm sure there won't be people rudely nitpicking and dismissing it in the comments for no real reason...
Par for the course on HN. You just gotta have thickskin and learn to acknowledge the comments without wholesale accepting them. Some nitpicks are great, some are flat out stupid. When it's your product you have the benefit of deciding which is which for your product (and accept the consequences, of course).
I don't get it. What is the difference between:
I want to pay less
I want to pay less, and here's a document an AI wrote for me
The analysis is generated via ~15 prompts and the output is pretty useful if you're going through buying a property with a mortgage.
The goal is to help mortgage borrowers get better deals on their mortgage. All mortgage terms are outlined in the loan estimate and closing disclosure documents as required by the CFPB (consumer financial protection bureau) - https://www.consumerfinance.gov/owning-a-home/loan-estimate
The tool requires uploading these documents since they have all the data needed to generate the analysis on the mortgage. Technically I can have the user paste in all the data via a textbox, but then it would be much more difficult to reliably parse. Also requiring the upload of the original documents is a much cleaner experience.
> The goal is to help mortgage borrowers get better deals on their mortgage.
I don't get it. Better, LendingTree, Credible, Bankrate, and a dozen other firms all aggregate mortgages. If there are some "best" mortgages in absolute terms, show them, you don't need anything on the form to do that; if there are offers that have lower upfront costs in exchange for higher rates, because the buyer is sensitive to upfront costs, show them. Why do you need to see the document in order to give the user what he wants?
Since you already know all of this, it begs the question: why do you specifically want to collect these documents? I'm not saying there is some conspiracy or anything.
The things you're mentioning allows you to compare mostly rates and points, but they don't do anything for origination or third party fees which can vary wildly from lender to lender. These fees MUST be disclosed in the loan estimate or the closing disclosure and its pretty much the only place you'll find them
Once I selected the offer, Better partner’s disclosure exactly matched the numbers shown in the comparisons. In other words the prices they quoted were accurate. Are you saying that they still are not?
Anyway, if you aggregated mortgage offers, you can do whatever you want. Make them as truthful and accurate as you want. Then you’d fulfill your goal, and you’d never need to see my documents. So why do you really need the documents? What is the angle?
Good job on the product and launching @aaln!
I can see it being useful for many home buyers.
I’m curious if AI tools like this are just wrappers around ChatGPT? Do they use their own LLM?
If you upload mortgage documents directly to ChatGPT can you get similar results?
Thank you @aoppaol!
AI tools like this don't generally need their own models or fine tuned LLMs. Uploading mortgage documents to ChatGPT may get some useful results but it really won't be comprehensive, it most probably won't take into account jurisdictional laws and mortgages rate data, and would be hard(er) to trust because of hallucinations without guardrails.
Technically this is an LLM wrapper but it utilizes around 5-10 prompts to extract and structure data from the document and ~15 prompts to analyze.
Hey. I ended up with an error saying my PDF could not be processed? But it’s my final loan disclosure form…
Also, I’m not entirely sure what costs this avoids? Inspection fees are already paid at this point, notary fees are paid to your broker, and transfer tax happens months later.
I ran out of anthropic credits, I just refilled and the service works well now
Most of the variance is in origination charges and title fees afaik. For example, borrower title insurance which is usually option can cost like 5k.
Not wanting to upload my info here, but what, concretely, does this tool do?
Does it scan to see if some of the fees I am agreeing to are higher than average? If I am paying for some services that should be no-cost or have no real value?
The tool analyzes mortgage loan estimates and closing disclosures to find every potential place a buyer might be overpaying, every negotiable fee, and show the buyer a comprehensive AI analysis so they don't get screwed over and can have peace of mind on buying their home.
When I was closing on a house, I called a few friends to help review my mortgage, and we found lots of mistakes. For example, I was getting charged transfer tax, which didn’t make sense for Florida where the seller typically pays that. The deeper I dug, the more I realized how much gray area there is with these documents - what’s negotiable, what’s inflated, what's normal in the property's jurisdiction, and what’s just non-competitive but seems ok since there's a lot of simplified complexity that goes into mortgages and what's an extra 1-3% on a mortgage that's "just going to be refinanced in 6 months" when interest rates go down.
Something that would be useful would be exporting to PDF with the whole report so you could send it to your realtor / lender, etc!
Thanks for mentioning this, I'm prioritizing this as a feature!
right now ive been trying to export the PDF but it's annoying
Going to work on this feature and ship it early next week if you're still in the closing process then.
This seems odd. If you need certain fees and line items explained to you, why not ask your lender? Surely you have a loan officer that can explain every item and why it is there.
If you don’t like some of the costs, go with another lender, or negotiate those with your loan officer.
I’m not sure what this tool is offering that individuals don’t already have.
This is supposed to be the role of the mortgage broker / loan officer.
However in practice it's more complicated. Homebuyers receive these documents after they already signed a binding contract to purchase. Most people don't even know what questions to ask, much less how to properly negotiate, and they're under the stress of just getting the deal done which include affording a down payment, getting a home appraisal, and making sure they're not getting into a bad deal.
Also throughout the closing, the real estate agent/mortgage broker/lender aren't incentivized to get the buyer the best deal; they usually just want to close. In general the mortgage broker/lender are not going to advise and how the customer how to shop around since it's technically illegal for them to give a "bad" loan to a buyer.
A set of mortgage documents can have over 100 pages and include 20+ fees. It can be difficult for a borrower/homebuyer to understand all these fees, and most have no idea if they're competitive. Loan officers or brokers aren't incentivized to get you the lowest rates on every line item. Sounds like this tool helps you understand if you're being overcharged.
why upload the actual document? why not let users enter only numbers?
Just checked out the tool it looks really interesting can definitely see the tremendous value to the standard consumer.
Just mining data. Please flag this.
this is great!!! Thanks so much for making this!
How are you handling the distribution?
hacker news, reddit, and other communities. Hopefully there will be a strong incentive in the future for real estate agents or mortgage brokers to recommend the tool. Also I think the name closing.wtf will be good for distribution.
I uploaded a mortgage I know well, and the results were pretty accurate.
Thanks, did you see any interesting insights you're comfortable sharing?
I was really just testing that the system works. I gave it a VA Home Loan and nothing seemed to be out of the ordinary in the analysis.
I wish I had this for payment processing.
here, take my data but tell me nothing about yourself.
It's a really new tool I quickly launched last week, there's a terms of service and privacy policy that gives info.
https://closing.wtf/terms-of-service https://closing.wtf/privacy-policy
I plan to go through SOC 2 / other compliancy certification in the near future. Happy to answer other questions
I use http://AimLoan.com! Every detail right is there for you to see including closing costs for whichever interest rate you decide to go with.
What's the PDF parsing like?
Extract all the text from the PDF, turn the pdf into images, send the text for each page along with the image to an LLM with a desired output strucutre.
You are not doing any of the fancy table extractor stuff?
i don't want to be the "get off my lawn" person or "old man yells at cloud," but i couldn't get to what the tool does because the landing page is so overwhelming with colors, the background, the borders, things moving around.
the CTA buttons across the site are inconsistent in shape, form, and color.
the text color in the footer has awful contrast, so it's hard to read.
Thanks for the feedback old man! I'm going to improve the aesthetic of the dark mode. You can try the light mode which should be easier to read right now.
>the text color in the footer has awful contrast, so it's hard to read.
Dark grey on white is awful contrast these days?
… I'm not them, but in dark mode, it's <50% grey on black, and for the headers, <32% grey on black. (#374151 on black, I think.)