📺 Watch the full tutorial on YouTube

💳 Needle 2 Finance Parser

Needle 2 Finance Parser is a completely offline, zero-latency expense tracking system that extracts structured financial records from messy text and voice notes using on-device AI. It updates your local financial records and refreshes visual budget dashboards instantly without sending sensitive financial data to cloud servers.

Step 1: Input	Step 2: AI Action	Step 3: Result
🗣️ Voice or raw text expense note	🧠 Needle 2 schema-constrained extraction	📊 Instant CSV update & dashboard chart refresh

---

🚀 Key Features

• 🔒 100% Offline Privacy: Process financial notes locally on your PC or mobile device without sending transaction details to external cloud services.
• ⚡ Zero-Latency Extraction: Converts unstructured text into structured fields in under 0.05 seconds using Needle 2's 14MB binary engine.
• 🎯 Guaranteed Data Shape: Enforces strict Pydantic schemas so transaction amounts, merchants, and categories are never malformed.
• 📈 Instant Dashboard Refresh: Automatically updates local CSV databases and renders real-time spending category charts.
• 🛡️ Confidence Scoring: Evaluates extraction accuracy locally and alerts users if spending categories need manual confirmation.

---

🛠️ Tech Stack

• AI Foundation Model: Needle 2 (cactus-needle 45M tool-calling model)
• Data Validation: Pydantic v2
• Programming Language: Python 3.10+
• Data Management: Pandas & CSV
• User Interface: Streamlit

---

📥 Installation & Setup

1. Install Dependencies

Open your terminal or PowerShell and run:

pip install cactus-needle pydantic pandas streamlit

2. Prepare Local Database

The system automatically initializes expenses.csv in your working directory upon first launch.

---

🏃 Running the Application

Launch the local Streamlit dashboard to input notes and view real-time charts:

streamlit run app.py

---

📁 File Structure

Needle-Finance-Parser/
├── app.py
├── expenses.csv
└── README.md

---

💼 5 Real-World Use Cases

1. 🛍️ Voice Expense Logging: Dictate quick purchase notes while driving or shopping without opening bulky banking apps.
2. 🧾 Receipt Summary Parsing: Copy text snippets from digital receipts or invoices directly into your local ledger.
3. 🚘 Travel Reimbursement Tracking: Capture per-diem meals and transportation expenses on offline mobile devices.
4. 🏠 Shared House Expense Ledger: Categorize utility bills and shared groceries locally with family members.
5. 📊 Private Budget Monitoring: Maintain complete privacy over personal cash flows without third-party subscription tracking.

---

🔮 5 Future Roadmap Features

1. 📱 Mobile Edge Integration: Deploy direct ARM binaries to Android and iOS local shortcut widgets.
2. 🎙️ Offline Whisper Audio Engine: Add local Whisper C++ voice transcription for hands-free offline audio notes.
3. 🔔 Custom Category Budget Alerts: Set automated spending thresholds per category with visual toast notifications.
4. 🏦 Bank Statement Reconciliation: Automatically cross-reference extracted cash notes against imported CSV statements.
5. 📤 Encrypted Export Formats: Support password-protected PDF and encrypted JSON financial report exports.

---

🔑 Keywords

needle2 needle-ai on-device-ai tool-calling offline-ai structured-extraction pydantic streamlit python-ai local-ai-parser edge-ai