ariadne-finance

name	ariadne-finance
description	Finance research add-on for Ariadne — PDF/Excel extraction, ticker recognition, financial knowledge graph.
version	0.10.0
author	kyssta
license	MIT
category	productivity
metadata	{"hermes":{"tags":["finance","research","pdf","tickers","knowledge-graph","ariadne-addon"],"requires":["ariadne-memory","ariadne-finance"]}}

Ariadne Finance — Finance Research Add-on

Finance research tools for the Ariadne memory system. Extracts financial data from PDFs, Excel, and CSV files, recognizes stock tickers, classifies market sectors, and builds a financial knowledge graph.

Installation

pip install ariadne-finance           # Excel + CSV
pip install "ariadne-finance[pdf]"    # + PDF support
pip install "ariadne-finance[full]"   # + yfinance market data

Tools

Document Ingestion

Tool	Purpose	Key Params
ariadne_finance_ingest	Ingest a financial document (PDF/Excel/CSV) into Ariadne	file_path, importance (0-1)
ariadne_finance_search	Search finance memories	query, limit, ticker, sector
ariadne_finance_tickers	Extract tickers from text or file	text or file_path

Entity Recognition

Tool	Purpose	Key Params
ariadne_finance_classify	Classify text into GICS market sectors	text
ariadne_finance_graph	Query financial knowledge graph	entity, hops

CLI Usage

# Ingest a financial report
ariadne finance ingest report.pdf --importance 0.8

# Search finance memories
ariadne finance search "NVDA revenue Q3"

# Extract tickers from a file
ariadne finance tickers report.txt

Dashboard API

# Ingest a document
curl -X POST http://localhost:8765/api/finance/ingest \
  -F "file=@report.pdf" -F "importance=0.8"

# Extract tickers from text
curl "http://localhost:8765/api/finance/tickers?text=AAPL+beat+MSFT+estimates"

# Classify sector
curl "http://localhost:8765/api/finance/classify?text=semiconductor+chip+manufacturing"

Python API

from arriadne_finance.extractors import PDFExtractor, recognize_tickers, classify_sector

# Extract from PDF
extractor = PDFExtractor()
result = extractor.extract("report.pdf")
print(result["content"][:200])      # extracted text
print(result["entities"])           # tickers, sectors found
print(result["tables"])             # extracted tables

# Recognize tickers in text
tickers = recognize_tickers("AAPL reported strong EPS, beating $MSFT")
# [{'type': 'ticker', 'value': 'AAPL', 'confidence': 0.7}, ...]

# Classify into sectors
sectors = classify_sector("semiconductor chip manufacturing GPU processor")
# {'Technology': 0.67, ...}

Workflows

Ingest and Search Financial Reports

# 1. Ingest reports
ariadne finance ingest q3-report.pdf --importance 0.8
ariadne finance ingest competitor-analysis.xlsx --importance 0.7

# 2. Search across all reports
ariadne finance search "revenue growth YoY"
ariadne finance search "NVDA" --k 5

Build Financial Knowledge Graph

from arriadne import AriadneMemory, AriadneConfig

config = AriadneConfig(db_path="finance.db")
mem = AriadneMemory(config=config)

# After ingesting reports, link entities
mem.add_edge("NVDA", "Technology", "belongs_to_sector", weight=1.0)
mem.add_edge("NVDA", "AMD", "competes_with", weight=0.8)
mem.add_edge("AAPL", "Technology", "belongs_to_sector", weight=1.0)

# Query the graph
result = mem.graph("NVDA", hops=2)

Entity Types

Entity	Description	Attributes
ticker	Stock ticker symbol	exchange, company_name
company	Company or corporation	sector, industry, country
sector	GICS market sector	—
financial_metric	Financial KPI	value, period
earnings_report	Quarterly/annual report	quarter, year

Tips

• Use importance=0.8+ for key reports, 0.5 for routine data
• The addon auto-discovers installed extractors — PDF requires optional deps
• Ticker recognition uses regex with false-positive filtering for common words
• Sector classification is keyword-based — works best with financial terminology