Mo

Mortgage Chat – A bot that answers your awkward home buying questions

Hacker News

Mortgage Chat – A bot that answers your awkward home buying questions

Mortgage Chat is a GPT-3.5-powered chatbot that answers your mortgage lending and home buying questions you were too afraid to ask. Working with mortgage loan officers can be challenging, off-putting, and frustrating. My goals for this project were: - Familiarizing myself with in-context learning using GPT3 - Exploring AI's potential in complex domains (non-technology) domains like mortgage lending - Creating a user-friendly, non-judgmental, and non-commercially motivated tool for 24x7 guidance on anything mortgage-related How I built it: - I built the bot using Langchain and used their sample chatbot as a base, although did significant prompt and other tweakings, including building the "next questions to ask" functionality. - I crawled the Consumer Finance Protection Bureau's website for mortgage content. Relevant content is fed to the bot from a FAISS index (Langchain makes this super easy). I used the CFPB as it struck me as the most neutral and factual of various online content sources. - I also provide the bot with up-to-date mortgage rate data from the Federal Reserve's FRED service. The bot could definitely be improved: - I tried letting GPT3 ask the user questions about their mortgage situation in order to be more helpful but found it overly aggressive and off-putting. I think this is definitely feasible, but requires more thought. - Websockets can be a pain, and I'd probably want to use server-side events in production to stream results to the bot client. h/t to: - Langchain and community. - The Consumer Finance Protection Bureau.

Share card

Actual performance

2points
1comments
Did not reach leaderboard

Launch Intel predictions

Analyze your own launch →
Indie HackersFits the IH revenue-focused audience · Strong signals: including · Missing: supports, reddit linkedin, podcasting
71%71% predicted probability of success on Indie Hackers, based on ML models trained on real launch data.
best fitHighest predicted score across all platforms for this description.
Product HuntOn track for Day 1 leaderboard · Strong signals: user, context, using · Missing: mac, agents, macos
71%71% predicted probability of success on Product Hunt, based on ML models trained on real launch data.
TrustMRRFits verified-revenue profile · Strong signals: answers · Missing: mobile apps, ios, personal
53%53% predicted probability of success on TrustMRR, based on ML models trained on real launch data.
AppSumoStrong fit for a featured deal · Strong signals: friendly · Missing: plus, platform, intuitive
52%52% predicted probability of success on AppSumo, based on ML models trained on real launch data.
Hacker NewsMay not resonate with HN audience · Strong signals: ide, io, including · Missing: https docs, excited, just released
46%46% predicted probability of success on Hacker News, based on ML models trained on real launch data.
nativeThis product was originally launched on this platform.
Acquire.comPre-revenue stage for this audience · Missing: arr, mrr, revenue
8%8% predicted probability of success on Acquire.com, based on ML models trained on real launch data.
BetaListMay not resonate with beta-testers · Strong signals: chat · Missing: web3, crypto, cryptocurrency
0%0% predicted probability of success on BetaList, based on ML models trained on real launch data.

Correct prediction on native model

Similar products

My
MySellerBot – Bot that answers 100s of questions without any training40%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

MySellerBot – Bot that answers 100s of questions without any training

Hacker News4
Br
Brisby, a Slack bot that answers questions you're sick of answering58%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Brisby, a Slack bot that answers questions you're sick of answering

Hacker News13
Br
Brren – Answers to Any Questions60%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Brren – Answers to Any Questions

Hacker News1
Deli Dahi - Bilgi Yarışması
Deli Dahi - Bilgi Yarışması79%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Crazy Questions. Brilliant Answers.

Indie Hackers1games
Po
Pokémon Go Chat Bot47%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Pokémon Go Chat Bot

Hacker News5
FleaMarketBot
FleaMarketBot43%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Twitch Chat Bot

Indie Hackers1$600/mobots
HumansMapped
HumansMapped66%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Answers to world questions

Product Hunt+13
Allen Bot
Allen Bot20%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Chat bot for tenants and landlords to communicate

Indie Hackers1ai
As
AskMe – answers repetitive support questions45%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

AskMe – answers repetitive support questions

Hacker News2
Qu
Quicktions – Answers to your quick questions48%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.

Quicktions – Answers to your quick questions

Hacker News42