About OpenLedger
Personal finance software built for freelancers and small business owners. AI-native, privacy-first, no subscription lock-in.
Our mission
Make personal finance software that respects your privacy, your wallet, and your time.
We believe managing your money should not cost money. The tools that freelancers, contractors, and small business owners need -- transaction tracking, categorization, tax reports, invoicing -- should be available to everyone, not locked behind $100+/year subscriptions.
OpenLedger is that tool. Free to download, free to use, and built with the same care and quality you would expect from premium software.
What we believe
-
Your data is yours.
No cloud storage by default. Crash reports for stability — analytics opt-in only. No selling your financial data.
-
Core features should be free.
Transaction tracking, categorization, reports, and invoicing. No paywalls.
-
AI should run on your device.
On-device ML and local LLMs by default. Cloud AI only with explicit opt-in.
-
Built for real people.
Not accountants. Not enterprises. Freelancers, side-hustlers, small business owners.
The story
OpenLedger started because we were frustrated. Quicken wanted $119/year to track our spending. QuickBooks wanted $180/year for features we would never use. Every "free" alternative was cloud-only, ad-supported, or a data-harvesting operation disguised as a finance app.
We wanted something simple: a desktop app that imports bank transactions, categorizes them intelligently, generates tax reports, and handles basic invoicing. No double-entry accounting. No enterprise complexity. Just the features a freelancer actually needs.
So we built it. OpenLedger is a native desktop application built with C# and Avalonia UI. It stores your data locally in SQLite. It uses on-device machine learning for transaction categorization. And it is completely free.
No venture capital. No growth-at-all-costs. Just good software, built with care.
Design philosophy
Simple over complex
"Money In/Out" not "Debits/Credits." Schedule C, not Trial Balance. If a freelancer would not understand it, we do not build it.
Local first
Your data lives on your computer. No cloud account required. No internet needed. Cloud features are optional add-ons.
AI-native
AI is not a bolt-on feature. Transaction categorization uses XGBoost ML + local LLM with tiered fallback. It learns from your corrections.
Quality engineering
Clean architecture. CQRS with MediatR. Domain-driven design. Integration tested. Built to last, not built to ship fast.
Technology
Desktop App
- C# / .NET 9
- Avalonia UI (cross-platform native)
- CommunityToolkit.Mvvm
- SQLite (local database)
- EF Core (data access)
AI / ML
- XGBoost ML model (on-device)
- Local LLM support (Ollama, LM Studio)
- Cloud AI opt-in (Gemini, Claude)
- Tiered fallback pipeline
- Web search enrichment for payee identification
Architecture
- Clean Architecture (4-layer)
- CQRS with MediatR
- Domain-Driven Design
- Repository + Unit of Work
- Integration tested