GithubHelp home page GithubHelp logo

begfi's Introduction

begfi Documentation

Website Logic (Non-Smart Contract & SQL)

User Onboarding (Borrower)

  1. Join Begfi:
    • User creates an account on Begfi.
    • Connects wallet.
    • Creates username.

Borrower Registration Process

Initial Steps:

  • Borrower gets to the website.
  • Before registration and username account creation, they are prompted to connect wallet via Sign In With Ethereum (SIWE): https://docs.login.xyz/
  1. Become a Borrower:
  • User chooses the "Borrower" option. The borrower button is on the top of the page.
    • New users are prompted with the question "Are you a Borrower" or are you a "Lender" and each box prompt has information like
    • "- This is a no-money down, no collateral, micro-loans platform.
    • First-time users can borrow up to $20 for a maximum of two months.
    • Verification via CIVIC (civic.me) is required to request a loan.
    • Loans can be repaid early, allowing for higher borrowing limits next time.
    • In your request, promise a repayment amount to the lender (anyone on the website).
    • Small loans, like $20, often have higher fees; expect to repay around $24 or more.
    • Choose an interest rate that attracts lenders while being affordable for you.
    • Timely repayment is crucial; late or unpaid loans are reported to lenders, affecting future borrowing ability." - This is followed by a prompt box explains what that entails and tells them they need to verify via CIVIC." -Lenders do not have to do anything or verify. They get a box prompt that explains they are welcome to start loaning on the website, what verification processes borrowers go through, and what rewards they will get.

Verification with Civic

  1. Verifies Wallet:

    • Borrowers connect their wallet to Begfi again to verify their CIVIC pass.

  2. Wallet Check:

    • Upon connecting the wallet, Begfi has a SQL entry to check if the wallet address is associated with a verified Civic identity in the users table (assuming a column civic_id exists).

    • SQL Query (Verification):

      SELECT civic_id
FROM users
WHERE wallet_address = ?;

  3. Verification Status:

Borrower On-boarding Verification Process:

Borrower Can See "Need to Borrow" button and Message Board borrow request buttons

  1. Form Display:
    • If BORROWER is verified:
      • Begfi displays "You can Request Loans" box prompt
      • and the Borrower Message Request Button (on the top of the landing page) goes from gray to orange.

Borrower Posts Request

  1. The Borrower Request button is a form to begin a Borrow Request follows a form where users can only enter certain information.

  2. In the SQL database, when user enters the following details for the loan request:

    • Amount requested ($AMOUNTREQUEST)
      • Amount of money being requested is a numerical value
    • Repayment time ($TIMEinDAYS)
      • Number of days until repayment is a numerical value of days
    1. Total amount to be repaid ($AMOUNT2PAY)
      • Amount to be paid back is a numerical value
    2. The reason why is a multiple choice option of different reasons in a dropdown.

  3. Below their profile UPON posting the request will be their loan information history (described later in the Loanbot info) so lenders can gauge the borrower's credit worthiness

  4. User History (Existing Borrowers):

  • Upon submitting the request, the system:
    • Fetches the borrower's history from the database (assuming tables like loans and borrower_history exist).
    • Calculates statistics:
      • Number of loans paid off (X total loans paid off, i.e. $1200 loans paid back)
      • Number of outstanding loans (Y amount of total loans outstanding, i.e. $50 loans outstanding)
      • Total number of loans (Z number of total loans. i.e. has received 4 loans)
    • Displays the calculated user history.

Loanbot Information Below Borrower Profile

Lender Response

  1. Lender Notification: Front-end website from UI/UX perspective:

    • The system identifies potential lenders and notifies them about the loan request.
      • Website can filter:
        • most recent borrower requests
        • borrower requests by size
        • first-time borrower requests (because these get higher rewards when a lender successfully lends to them)
        • this allows lenders to also scroll through and find people they want to lend to
    • The lender can also see in the message board new borrower requests by just scrolling

  2. Lender Decision:

  • Lender looks at a borrower's request
  • Once again the request looks like BORROWER needs to borrow "X" amount of money for "Y" days and will payback "Y" amount because of reason (from a multiple choice dropdown) "Z" reason.
  • There is a button to send borrower the loan under each username request on the message board called "Send Loan"
  1. What is the Process after Lender accepts request and sends money?
    • Lender hits the button "Send Loan"
      • This creates a prompt from Metamask to send that EXACT amount
    • After money is sent a box prompt tells Lender they have loaned an amount of money to the borrower, i.e. "Nice job! You have lent "X" username money"
    • Additional information can include:
      • Once the loan is paid off how many rewards the lender will receive
      • Lenders receive an additional 25 Beg tokens for a first-time borrower. Lenders receive 20 Beg tokens for second time-borrowers. Third-time borrowers means lenders get 15 Beg tokens.
      • Calculation follows:
        • For every dollar of a loan, users receive one Beg token, where 1 Beg token equals 1 dollar. For example, a $100 loan results in 100 Beg tokens.
        • For all subsequent loans after the third loan the lender makes, the reward is a flat rate of 10 Beg tokens.
          • Prompt if they need a notification once the loan is repaid # Lender can put a messaging service like Telegram to inform them when loan is due.
          • Issue: lenders may not want to do notifications though. They're busy people. Likely, no one uses the notifications option.

Lender Sends Loan (Smart Contract & Blockchain)

  1. Lender Send Loans:
  1. After the Query, and the txn ID is confirmed to have gone to the right person, the SQL database is updated: - USERNAME ID of borrower - USERNAME ID of lender - Since the query has checked if the loan was the correct amount according to what the Lender requested, the SQL database updates: - “Loan Sent"
    - "Not paid"
    - “Paid" - The front-end website also html changes the text in a box next to the borrower to the correct terms "- “Loan Sent" - "Not paid" - “Paid""

More info on Transaction check via The Graph

  1. Blockchain Query:
  • Begfi uses the blockchain's API to query the transaction and confirm if the funds ($AMOUNTREQUEST) have been transferred to the smart contract.
  • This query occurs across all loans every 30 seconds.
  • The cost of this for 5000 loans PER MONTH is $71 per year since The Graph has 100,000 free monthly queries
    • To determine the total monthly cost of querying 5,000 loans every 30 minutes, with an average loan length of 7 days, we start by calculating:
      • the number of queries per loan.
        • Each loan is queried every 30 minutes, resulting in 2 queries per hour, 24 hours a day, for 7 days, yielding 2×24×7=3362 \times 24 \times 7 = 3362×24×7=336 queries per loan.
        • For 5,000 loans, this totals 336×5,000=1,680,000336 \times 5,000 = 1,680,000
        • 336×5,000=1,680,000 queries in a week. Extending this to a month (approximating 30 days), we get 1,680,000×307≈7,200,0001,680,000 \times \frac{30}{7} \approx 7,200,0001,680,000×730​≈7,200,000 queries.
        • With 100,000 free queries each month, the chargeable queries amount to 7,200,000−100,000=7,100,000
        • At a cost of $0.00001 per query, the total monthly cost is 7,100,000×0.00001=$717,100,000 \times 0.00001 = $717,100,000×0.00001=$71.
        • Therefore, the total cost per month for querying 5,000 loans every 30 minutes is approximately $71.
        • This number would not change if it was 7 days, 30 days, or more for the average loan length
  • If such a situation where 1) total sum does not equal $75 (where the borrower has not paid in full) or has paid in batches so that you just look at all the transactions from person A to person B, between the time when the loan was given and the loan is due, and add the numbers up.
  • Create the query, in The Graph, https://www.youtube.com/watch?v=EJ2em_QkQWU (doc video explaining how to create query) whereby it checks every 3 minutes if BORROWER has repaid the loan to the Lender.
  1. SQL Entry:
  • Upon successful transfer confirmation:
    • Update the loan status in the database (e.g., loans table)
    • Record the loan details in the loans table with relevant columns like borrower_id, lender_id, amount, time_to_repay, total_amount_to_pay, and status.
    • Front-end updates to indicate the wallet has received the borrower's transfer from the borrower's wallet. If the transfer is in small batches, it triggers the front-end to update to: - "Loan Sent" - "Not Paid" - "Paid"

The Borrower Repays Loan / Lender is Paid Back and Potential Post-MVP Messaging System from Borrower to the Team

  1. Borrower sends back money.
    • Prompt for sending it back exists called "Pay Back Lender"
    • After money has been sent by the borrower then a notification at the end of the payment date via The Graph, API fetch if loan has been paid back by showing if address of borrower has sent money to address of lender
  2. The query sums up and equals the correct amount and therefore creates a trigger to make sure the SQL database changes.
    • This is done through a microservice like calculation in The Graph
  3. To ensure this process works correctly, the borrower must not directly send money to the lender’s address by copying it and initiating a transfer. While this method technically works and The Graph will still receive accurate data, the BORROWER must repay the loan via the designated button. This is crucial because neither the lender nor the borrower should see each other’s addresses on the message board, only their usernames. This prevents several issues: - The borrower might accidentally find and use the lender’s address with a different wallet. - Even though the borrower can see the address in the transaction, we assume that newer borrowers may not check the address in their wallet. - Borrowers might mistakenly think they need to copy the address for repayment. - A borrower’s friend could use the address and send money on the borrower’s behalf without the lender knowing.
  4. Finally, post-MVP, in the rare case that there’s an issue from the borrower’s side, they can contact the Begfi team through a web3 messaging system widget available exclusively to borrowers: dm3 network. The mission of this system is to:
    • "Build with dm3: Add web3 messaging to your dApp. The dm3 protocol simplifies adding messaging capabilities to a dApp. With just a few lines of code, you can integrate our customizable widget. This widget can be used for support chat, in-app communication, or any other type of messaging within the dApp. Whether you want to offer support, facilitate social interactions, or create any other type of in-app communication, the dm3 protocol can help you do it easily and quickly."

Loan Bot to show Loan Information to Lender and its Functionality##

Basic reddit Loanbot documentation:

  • https://www.reddit.com/r/LoansBot/comments/2ea912/loansbot_basic_usage/

  • https://botwiki.org/bot/loansbot/

  • Bot should probably work by taking the total amount owed instead of the initial amount.

    • For example, if the bot would take only $50 but the lender gets repaid $75 USD, then it might be confusing. According to the GitHub documentation of /loanbot,
      • "The LoansBot does not consider interest:
        • A loan is considered repaid when the entire principal has been repaid, regardless of any outstanding interest. This is also reflected on the ban status of users (in most circumstances).
        • So a $100 loan is considered fully repaid when the borrower has returned $100."
      • Another way of looking at this is that interest amounts are not considered part of the info relevant to the LoansBot.

  • All loans are in USD. This is not because we discriminate against other currencies, but almost all important stablecoin transactions occur in USD-related stablecoins.

Reasons Why Begfi Should Succeed

Offering small amounts of credit to individuals who may not have access to traditional banking. However, the traditional finance (TradFi) space often falls short in terms of accessibility, speed, and cost-effectiveness. Begfi, leveraging the power of blockchain technology and Web3 verification, aims to revolutionize microloans. This document outlines why crypto-based microloans are superior, the importance of Web3 verification, and the potenti

A. Advantages Over Traditional Platforms:

  • Unlike LendingClub (down over 95% in stock price), LendingTree, and Prosper, which are all platforms where you invest and they issue loans, these services don't involve peer-to-peer lending. Instead, they act as intermediaries for your transactions.
  • (Venmo/CashApp/ApplePay/PayPal) are more willing to shut down accounts they think are being used to do p2p loans.

B. Market Trends:

  • Compared to before, there is a larger base of customers for crypto-based loans worldwide.
  • In the developing world, higher interest rates mean higher bank loan requirement restrictions than ever (due to higher interest rates) that hurt low-income people from getting personal loans.
  • APRs for p2p lending are still very good compared to bonds in MOST countries (a 10% APR arbitrage opportunity).
  • APRs for p2p loans are still LOWER than most bank payday loans.

C. Marketing Potential:

  • It is Tiktokable. An influencer can be paid $500 to talk to their 100,000 followers to go to Begfi instead of a bank because bank loans are now 9% APR with a mortgage/house as collateral, no-collateral loans are 17%+, but they can still freeze your bank account and go after your assets by re-selling the loan to debt collection agencies.

Code References

Key Points:

  1. Loan Details:

    • Loan amount: Specified in USDT.
    • Expected repayment: Specified amount in USDT (e.g., the borrower asks for 50 USD and is expected to repay 75 USD in 30 days).
    • Repayment in batches over different days is possible.

  2. Service Requirements:

    • Periodically check (every 5 minutes) if the borrower has sent money to the lender.
    • Calculate the total amount repaid by the borrower.
    • Verify if the total repayment meets or exceeds the expected amount.
    • Determine if the repayment deadline has passed.

Provided Solution:

  1. Environment Setup:

    • Use Node.js with web3.js to interact with the Polygon network.
    • Use axios to fetch transaction data from the PolygonScan API.

  2. Script Details:

    • Initialize Web3 and connect to the Polygon network.
    • Set up a polling mechanism to check for transactions every 5 minutes.
    • Filter transactions to see if payments have been made and calculate the total amount paid.
    • Check if the total repayment meets or exceeds the expected amount and determine if the repayment deadline has passed.

To call MATIC RPC:


const Web3 = require('web3');
const axios = require('axios');

// Connect to the Polygon network
const web3 = new Web3('https://polygon-rpc.com/');

// Addresses and loan details
const lenderAddress = '0xLenderAddress';
const borrowerAddress = '0xBorrowerAddress';
const expectedTotalRepayment = 75; // in USDT
const repaymentDeadline = new Date('2024-08-01'); // 30 days from the loan start

// USDT contract address on Polygon
const usdtAddress = '0x3813e82e6f7098b9583FC0F33a962D02018B6803'; // USDT contract address

// Function to check if repayment is complete
const checkRepayment = async () => {
    let totalRepaid = 0;

    try {
        // Get the transaction history for the borrower address
        const response = await axios.get(`https://api.polygonscan.com/api`, {
            params: {
                module: 'account',
                action: 'tokentx',
                address: borrowerAddress,
                startblock: 0,
                endblock: 'latest',
                sort: 'asc',
                apikey: 'YourPolygonScanAPIKey'
            }
        });

        const transactions = response.data.result;

        // Filter transactions to the lender address involving USDT
        const usdtTransactions = transactions.filter(tx => 
            tx.to.toLowerCase() === lenderAddress.toLowerCase() &&
            tx.contractAddress.toLowerCase() === usdtAddress.toLowerCase()
        );

        // Calculate the total amount repaid by the borrower
        usdtTransactions.forEach(tx => {
            const valueInUsdt = web3.utils.fromWei(tx.value, 'mwei'); // USDT uses 6 decimals
            totalRepaid += parseFloat(valueInUsdt);
        });

        console.log(`Total repaid: $${totalRepaid.toFixed(2)} USDT`);

        // Check if the total repaid amount meets or exceeds the expected repayment
        if (totalRepaid >= expectedTotalRepayment) {
            console.log('Loan has been fully repaid.');
        } else if (new Date() > repaymentDeadline) {
            console.log('Loan repayment deadline has passed.');
        } else {
            console.log('Loan is still being repaid.');
        }
    } catch (error) {
        console.error('Error fetching transactions:', error);
    }
};

// Polling mechanism to check every 5 minutes
setInterval(checkRepayment, 5 * 60 * 1000);

// Initial check
checkRepayment();

begfi's People

Contributors

snak2e-tech avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.