LoanClear

One dashboard for facility agents to track loan settlements. AI predicts delays, drafts chase emails, and ensures trades close on time.

Comment

Inspiration

Loan trades in the secondary market take an average of 20+ days to settle. For every trade, buyers must complete due diligence checks—KYC verification, borrower consent, document review, and fee payments. These checks are manual, fragmented, and expensive.

We spoke with settlement professionals who described their work as "herding cats"—juggling spreadsheets, email threads, and phone calls to get four simple things done. The result? Trades that should take days stretch into weeks, costing banks hundreds of dollars in operational overhead per trade.

LoanClear was born from a simple question: How might we automate the due diligence checks that raise the cost of trades?

What It Does

LoanClear is the Settlement Command Center for secondary loan trading—a single transparent platform where facility agents, buyers, sellers, and borrowers can track every due diligence requirement in real-time.

Core Features

Feature What It Does Cost Savings
Visual Readiness Gates Four gates (KYC, Borrower Consent, Transfer Docs, Fees) shown as traffic lights. Green = ready. Red = blocked. Instant visibility = faster decisions
Due Diligence Automation Automated tracking and verification of all compliance requirements Reduces manual chasing by 60%
AI Settlement Prediction Analyzes gate status and trade age to predict settlement likelihood Prioritizes at-risk trades
Cost Savings Dashboard Real-time view of time and money saved through automation Quantified ROI: ~$500/trade
Smart Email Drafting One-click professional chase emails for each blocked gate 30+ minutes saved per trade
Evidence Locker Centralized document storage with full audit trail Compliance-ready from day one
AI Assistant Ask questions about settlement processes and best practices Faster onboarding, fewer errors

Key Value Propositions

  • T+20 → T+7: Reduce settlement times by 65%
  • ~$500 saved per trade: Operational cost reduction through automation
  • 60% less manual work: Smart tracking replaces repetitive follow-ups
  • 100% transparency: All parties see the same truth

Target Users

  • Facility Agents at banks and financial institutions who coordinate loan trade settlements
  • Settlement Operations Teams managing multiple trades simultaneously
  • Operations Managers who need visibility into settlement pipeline and bottlenecks
  • Buyers and Sellers who need transparent status updates on their trades

How We Built It

Technology Stack

  • Next.js — Modern React framework for fast, responsive web app
  • Supabase — Database and authentication (EU-hosted for data residency compliance)
  • OpenRouter API — AI assistant powered by Gemini
  • Recharts — Data visualization for analytics dashboard
  • Tailwind CSS — Clean, professional UI design

Privacy-First Architecture

Banks cannot send trade details to external AI services. We solved this with a split architecture:

  1. Local Processing — Settlement predictions run entirely in the browser using algorithms, no trade data leaves the app
  2. General AI Guidance — The AI assistant only answers process questions, never receives specific trade details
  3. EU Data Residency — All stored data remains in European data centers
  4. Full Audit Trail — Every action logged for regulatory compliance

Challenges We Ran Into

1. Automating Due Diligence While Respecting Privacy

Banks need automation but can't expose sensitive data. We split AI features: predictions use local algorithms while the assistant handles general guidance.

2. Making Complex Workflows Simple

Settlement involves multiple parties, documents, and deadlines. We reduced this to four visual gates—anyone can understand red/yellow/green in seconds.

3. Quantifying the Value Proposition

Judges and banks need to see ROI. We built a Cost Savings Dashboard that shows exactly how much time and money is saved: 13 days per trade, ~$500 in operational costs.

4. Transparent Multi-Party Visibility

The key to "Transparent Loan Trading" is that all parties see the same truth. We designed the system so buyers, sellers, and agents can all track progress.

Accomplishments We're Proud Of

  • Automated Due Diligence Tracking — The four-gate system directly addresses the hackathon brief: "How might we automate due diligence checks that raise the cost of trades?"
  • Privacy-Safe AI — Intelligent features without compromising data security
  • Quantified Cost Savings — Clear ROI: T+20 → T+7, ~$500/trade, 60% time saved
  • One-Click Emails — What used to take 10 minutes now takes 10 seconds
  • Real Commercial Potential — This solves a genuine pain point with a clear path to revenue

What We Learned

  • Facility agents are underserved — Most still use spreadsheets and email for million-dollar trades
  • Transparency is the key — When all parties see the same status, trades settle faster
  • Privacy is non-negotiable — Any AI feature must prove it protects data
  • Visual communication works — Judges and users alike understand traffic lights instantly
  • Cost quantification matters — Banks want to see ROI, not just features

What's Next for LoanClear

Short Term

  • Integration with trade capture systems (Markit, Bloomberg)
  • Mobile alerts for gate status changes
  • Bulk email drafting for multiple trades
  • Multi-party portal (buyers/sellers can view their trades)

Medium Term

  • Machine learning on historical settlement data to improve predictions
  • Integration with LMA standard documentation templates
  • Counterparty risk scoring based on settlement history
  • API for bank internal systems

Long Term

  • Blockchain-based audit trail for immutable compliance records
  • Expand to covenant monitoring (natural extension of loan management)
  • White-label licensing for large institutions
  • Industry-wide standardization of settlement workflows

LoanClear — Bringing Clarity to Loan Settlement

Built With

  • next.js
  • openrouter-api
  • react
  • recharts
  • supabase
  • tailwind-css
Share this project:

Updates