Skip to content

Pure Python Pipeline Overview

Introduction

The Pure Python Pipeline is a high-performance, deterministic alternative to AI-based portfolio analysis that delivers 10-20x speed improvements and 100% cost reduction while maintaining analysis quality.

Key Benefits

Performance

  • Speed: 10-20x faster than AI-based analysis
  • Cost: $0.00 per analysis (zero LLM calls)
  • Consistency: 100% deterministic results
  • Scalability: Handles large portfolios efficiently

Quality

  • Accuracy: Maintains analysis quality
  • Reproducibility: Same inputs always produce same outputs
  • Testability: Easy to unit test and validate
  • Auditability: Transparent algorithms

Architecture Components

The pipeline consists of four integrated components:

1. Portfolio Deep Analyzer

Replaces AI-based DeepAnalysisCrew with fast, deterministic Python calculations.

  • Pure Python scoring using DeepAnalysisScorer
  • Real market data fetching via QuantitativeAnalysisTool
  • JSON export generation for downstream systems
  • HTML report generation using Jinja2 templates

Performance: < 1 second per holding, 0 LLM calls, $0.00 cost

Learn more →

2. A+ Discovery Integrator

Integrates A+ discovery with deep analysis results by reading JSON exports.

  • Scans output directories for analysis JSON files
  • Identifies A+ and A grade holdings
  • Consolidates opportunities across asset classes
  • Fixes "0 opportunities found" issue

Learn more →

3. Backtesting Pipeline Connector

Automatically executes backtesting when A+ candidates are available.

  • Reads A+ candidates from discovery results
  • Executes backtesting for each candidate
  • Generates performance metrics
  • Saves results to JSON for report integration

Learn more →

4. Python Report Generator

Generates comprehensive HTML reports using Jinja2 templates (no AI).

  • Template-based HTML generation
  • Portfolio statistics calculation
  • Holdings analysis with grades and scores
  • Deep analysis integration
  • Performance metrics display

Learn more →

Data Flow

Text Only
Portfolio Holdings
    ↓
[Deep Analysis] → JSON exports (output/{asset_class}/*.json)
    ↓
[A+ Discovery] → Identifies A+ opportunities
    ↓
[Backtesting] → Performance metrics for A+ candidates
    ↓
[Report Generation] → Final HTML report

View detailed data flow →

Performance Comparison

Speed

Component AI-Based Python-Based Improvement
Deep Analysis (per holding) 30-60s <1s 30-60x faster
A+ Discovery 20-40s <1s 20-40x faster
Backtesting 10-20s 2-5s 2-4x faster
Report Generation 30-60s <1s 30-60x faster
Total (5 holdings) 5-10 min 10-20s 15-30x faster

Cost

Component AI-Based Python-Based Savings
Deep Analysis (per holding) $0.05-0.10 $0.00 100%
A+ Discovery $0.02-0.05 $0.00 100%
Backtesting $0.01-0.02 $0.00 100%
Report Generation $0.05-0.10 $0.00 100%
Total (5 holdings) $0.65-1.35 $0.00 100%

View detailed performance metrics →

Quick Start

Python
from finwiz.scoring.portfolio_deep_analyzer import analyze_portfolio_with_python
from finwiz.integration.aplus_discovery_integrator import integrate_aplus_discovery_with_deep_analysis
from finwiz.integration.backtesting_pipeline_connector import connect_backtesting_to_discovery_results
from finwiz.reporting.python_report_generator import generate_python_report

# Run complete pipeline
analysis_results = analyze_portfolio_with_python(holdings, session_id)
discovery_results = integrate_aplus_discovery_with_deep_analysis(session_id)
backtesting_results = connect_backtesting_to_discovery_results(session_id)
report_path = generate_python_report(portfolio_review, analysis_results, session_id)

View complete usage guide →

Documentation Structure

Next Steps

  1. Learn about components - Understand each pipeline component
  2. Follow the how-to guide - Implement the pipeline
  3. Review best practices - Optimize your implementation