Nixora — Digitising the Loan Market

Inspiration

The global loan market is worth trillions of dollars, yet it still runs on PDFs, emails, and manually maintained spreadsheets.
Even the most sophisticated lenders and agents rely on people to read, interpret, track, and re-enter data from long legal documents.

Behind every major business — from airlines to energy companies — sits a loan.
And behind every loan still sits a PDF.

This creates three systemic problems:

Risk is hidden inside unstructured text
Liquidity is slow because loans cannot be easily compared or traded
Operations are expensive because every lifecycle event is manually tracked

Loan documents today exist as legal prose — opaque, unsearchable, and disconnected from the systems that actually manage money, risk, and compliance.

Nixora changes that.

The Problem

A modern commercial loan may last 5–10 years and include:

Multiple lenders
Variable interest rates
Dozens of covenants
Continuous reporting and payment obligations

Yet all of this is buried inside documents that cannot be queried, analysed, or automated.

So institutions are forced to rely on:

Manual compliance checks
Spreadsheets to track maturities
Emails to coordinate between teams
Legal reviews just to know what a loan requires

This leads to:

Missed obligations
Late payments
Compliance failures
Hidden portfolio risk

The financial system is digital,
but the infrastructure behind loans is still analog.

Our Solution

Nixora turns loan agreements into live digital assets.

By combining Google Document AI, Vertex AI, and Gemini 3.0 pro, Nixora ingests raw credit agreements and converts them into LMA-standardised, structured data.

Every borrower, lender, covenant, maturity date, and obligation becomes:

Searchable
Comparable
Trackable
Actionable in real time

And with Nixora’s AI-powered Loan Chat, users can now ask natural-language questions like:

“What is the availability period for this loan?”
“Which covenants apply to this borrower?”
“When is the next interest payment due?”

…and receive instant, accurate answers directly from the agreement.

Instead of managing loans as files, Nixora lets institutions manage them as data — interoperable, auditable, and connected across systems.

What it does

Nixora is an end-to-end Digital Loan Operating System for the global loan market.

Users can upload any loan agreement (PDF, Word, Excel), or capture documents directly from the web using the Nixora Chrome Extension.
From there, Nixora transforms unstructured legal contracts into live, intelligent, and standardized loan data.

Nixora then:

Extracts and normalizes text using Google Document AI
Understands and structures the contract using Vertex AI + Gemini 3.0 Flash Preview
Maps everything into LMA-aligned JSON — creating a real-time, structured loan model

From a single upload, Nixora powers the entire loan lifecycle:

Portfolio Search & Querying
Instantly search across borrowers, lenders, facilities, covenants, currencies, margins, and maturities across your entire book of loans.
Loan Comparison
Compare pricing, structure, repayment terms, and risk across multiple facilities and borrowers in seconds.
AI Loan Chat
Ask natural language questions about any agreement — Gemini returns precise, explainable answers grounded directly in the contract.
Risk, Covenant & Maturity Reporting
Automatically detect high-risk loans, upcoming deadlines, covenant exposure, and compliance gaps.
PDF & Excel Exports
Generate institution-grade reports for credit committees, auditors, and regulators.
Loan Schedules & Event Tracking
Automatically create payment schedules, covenant tests, and maturity timelines from the legal text.
Calendar & Workflow Sync
Push obligations directly into Google Calendar and Trello to ensure nothing is missed.
Real-time Collaboration
Invite colleagues via WhatsApp to view, edit, and work on the same loan live.
Push Notifications
Desktop alerts for uploads, payments, covenant breaches, and critical risk events.
Chrome Extension for Live Loan Capture
Capture loan documents, term sheets, and deal data directly from emails, data rooms, and web portals — sending them instantly into Nixora for structuring and tracking.

Nixora turns static loan documents into living financial systems that teams can search, monitor, collaborate on, and trust.

Why this matters commercially

Value Proposition

Nixora gives financial institutions something they have never had:

A single, intelligent source of truth for every loan.

From executives to analysts, everyone sees the same live data, the same risks, and the same obligations — without digging through documents.

Efficiency Gains

What currently takes teams of analysts, lawyers, and operations staff becomes:

Automated
Real-time
Always accurate

Nixora eliminates manual document review, spreadsheet tracking, and missed deadlines — dramatically reducing operational cost and risk.

Scalability

Once a loan is digitised:

It can be searched
Compared
Queried
Traded
Analysed

Nixora scales like software, whether managing 10 loans or 10 million.

Market Impact

By turning legal contracts into digital assets, Nixora unlocks:

Faster loan trading
Better risk pricing
More transparent credit markets
Easier access to capital for businesses

Nixora is not just a tool — it is financial infrastructure.

How we built it

Nixora is a cloud-native, API-driven platform.

AI & Document Intelligence

Google Document AI – text extraction
Google Vertex AI + Gemini-3.0-pro – clause understanding, field extraction, LMA mapping, and AI loan chat

Backend

Java + Spring Boot
PostgreSQL (JSONB loan model)
Redis for fast querying
Cloudinary + Google Cloud Storage for documents

Integrations

Google Calendar API – obligation syncing
Trello API – workflow management
Firebase Cloud Messaging – push notifications
Resend – transactional email

Frontend

React + Next.js
Chrome Extension for in-workflow loan access

All functionality is exposed via a fully documented OpenAPI / Swagger API.

API Surface

Core

POST /api/loans/documents/upload
PATCH /api/loans/documents/update-field/{loanId}
GET /api/loans/documents/getLoan/{loanId}
GET /api/loans/documents/getAllLoan
DELETE /api/loans/documents/delete-loan/{loanId}

Portfolio Intelligence

POST /api/loans/query/search-query
POST /api/loans/compare
GET /api/reports/portfolio
GET /api/reports/high-risk
GET /api/reports/maturity

AI Loan Chat

POST /api/loan-chat/chat/{loanId}

Exports

GET /api/reports/export/*/pdf
GET /api/reports/export/*/excel

Scheduling & Workflow

GET /api/loans/{loanId}/schedule
POST /api/loan/calendar/{loanId}/sync
POST /api/trello/connect
POST /api/trello/push-field

Collaboration

POST /api/collaboration/invite
POST /api/collaboration/accept

Notifications

POST /api/notifications/register-device

Nixora System Architecture

Challenges We Encountered

Accurate data extraction: Early attempts using regular expressions were unreliable for complex legal language. We solved this by combining Google Document AI, Vertex AI, and Gemini 3.0 Flash, which enabled us to extract and structure loan data far more accurately.
LMA standardization: Identifying true LMA-standard fields was challenging, as there is no fully open public schema. We conducted deep research using LMA publications, sample agreements, and industry articles, allowing us to build a high-quality standardized model that closely reflects real-world LMA structures.
Access to real loan agreements: Obtaining real-world commercial loan documents was difficult due to confidentiality. Through extensive public research and financial document repositories, we sourced realistic agreements to properly test and validate Nixora’s AI pipeline.

Accomplishments we’re proud of

True digital loans — not PDFs, but live data
LMA-aligned structured loan models
AI-powered loan chat for instant answers
Calendar, Trello, and push notification integrations
Real-time collaborative loan editing
Institutional-grade reporting & exports

What we learned

Loan professionals don’t want more tools — they want their existing tools to become smarter.
AI is most powerful when paired with structured data.
Integrations drive adoption as much as intelligence.